Hoffman Amplifiers Tube Amplifier Forum
Amp Stuff => Tube Amp Building - Tweaks - Repairs => Topic started by: pullshocks on October 13, 2024, 03:41:59 am
-
Recently I came across some information about bias, plate voltage, cathode current, screen voltage, screen current, dissipation, power supply dropping resistors, etc with the various versions of the Fender Champ. D-lab Electronics makes a Fender Vibro Champ sound better than new | The Gear Page (https://www.thegearpage.net/board/index.php?threads/d-lab-electronics-makes-a-fender-vibro-champ-sound-better-than-new.2343139/) It points out the 5F1 had screen voltage of 295 V, compared to plate voltage of 340 V. thanks to a 10K power supply dropping resistor.
The RCA Tube Manual 6V6 Characteristics and Typical Operation seems to line up with that information.
This got me wondering about these issues in my own mashup 6V6 SE amp. It has a 1K screen resistor but it is not dropping screen voltage very much, only a few volts.
On the attached schematic/block diagram you can see the numbers I get with a 5Y3 and with a 5V4. Especially with the 5V4, it looks like the screen current goes above the maximums during loud/hard/fast Ramones style strumming.
Am I right in thinking I need to increase the screen resistor substantially?
-
If the choke can cope with about 60mA, I suggest to move swap it to be between the A and B nodes. Try various values 1k to 10k between the B and C nodes to achieve the desired 6V6 conditions.
-
Overall, I would be content with your amp as-is, based on what you've shared already.
... Especially with the 5V4, it looks like the screen current goes above the maximums ...
The line-items you drew the red box aren't aren't "maximums" in the form of "a value that must not be exceeded." Instead, they are the values achieved in the data sheet condition when maximum power output is occurring with signal applied.
Recently I came across some information about bias, plate voltage, cathode current, screen voltage, screen current, dissipation, power supply dropping resistors, etc with the various versions of the Fender Champ. D-lab Electronics makes a Fender Vibro Champ sound better than new | The Gear Page (https://www.thegearpage.net/board/index.php?threads/d-lab-electronics-makes-a-fender-vibro-champ-sound-better-than-new.2343139/) It points out the 5F1 had screen voltage of 295 V, compared to plate voltage of 340 V. thanks to a 10K power supply dropping resistor.
The RCA Tube Manual 6V6 Characteristics and Typical Operation seems to line up with that information.
This got me wondering about these issues in my own mashup 6V6 SE amp. ...
The voltages & dissipations you posted for your amp are plenty-fine for the 6V6, even with the 5V4. Yes, the rating system used in the data sheet results in a plate dissipation rating of "12 watts" but a later rating system said "14 watts" for the same tube.
Further, the vintage Fender amps being discussed in that TGP thread have higher voltages & dissipations at all levels, hence the D-Lab video about fixing the Vibro Champ. Except it's only nerds like us that see "a problem": the amp's owners used it for 5 decades without an apparent issue.
Lastly, your screen current might not be as high as you think, depending on how you're measuring it. If you're using a modern DMM you might be seeing an RMS-value (sine-wave peak x 0.7071) where RCA in 1937 (whose lab tests generated the data sheet values (https://www.one-electron.com/Archives/RCA/RCA-AppNotes/RCA%201937%20AN-80%20Operation%20of%20the%206V6-G.pdf)) were probably using a meter with an "average-responding meter (https://www.fluke.com/en-us/learn/blog/electrical/what-is-true-rms?srsltid=AfmBOortrqlfgL8ybo013ZS5a0Mx_EwFKvLJBE7iu2XFpikwSxtMun-S)" (sine wave peak x 0.637). This latter type of meter accurately measures pure sine waves, but screen current is never a pure sine wave. It's very low over most of the signal cycle, then has a brief, relatively high-current blip at maximum plate current (or not, depending on the plate load used).
-
What you've built is similar to the 5E1 Champ output stage. Fender later dropped the screen voltage for the 5F1, and I do wonder if that tells us anything about screen reliability.
To add to what HotBluesPlate said, the easy way to measure average screen current is to measure the DC drop across the screen stopper.
-
What you've built is similar to the 5E1 Champ output stage. Fender later dropped the screen voltage for the 5F1, and I do wonder if that tells us anything about screen reliability.
To add to what HotBluesPlate said, the easy way to measure average screen current is to measure the DC drop across the screen stopper.
That is how I measured screen current. Conveniently, the 1K resistor measured at 1003 ohms, so no math needed. My DMM is an Extech "True RMS." With my Johnny Ramone stress test (banging away at an open A chord as fast and evenly as I can) the display fluctuates rapidly. 9 mV was the highest digit I could read.
Speaking of measurement techniques, can a DMM be used to measure signal strength coming into the grid?
-
9 mV was the highest digit I could read.
I assume you mean 9V dc. So it was more of a rare peak reading rather than the average?
Speaking of measurement techniques, can a DMM be used to measure signal strength coming into the grid? 9mA *average* does seem rather high for a 6V6
I trust a DMM to measure a sine wave between 50-100Hz. Anything else is a 'maybe'.
-
can a DMM be used to measure signal strength coming into the grid?
I look at probing grids like Schrödinger looks at cats in a box
you can find valid data, you can fin erroneous data, or you can cause unintended consequences
a real handy feature on my Fluke is the min/max/avg you can monitor while wailing away for 15 minutes, then kick back and do some easy math, tweak if necessary then play like you stole it
-
I really appreciate all the responses.
I increased the screen resistor to 9K (added an 8K in series, the biggest 2 watter in my stash).
1K 9K
Plate 326 v 327 v
Screen 323 v 299 v
cathode 20 v 18.9 v
Cath resistor 458R 458R
Cath current 43.6 mA 41 mA
Dissipation 13 W 12.6 W
Scr. current max 9 mA 4.6 mA
It does not sound any better to me with the larger screen resistance. Maybe worse
EDIT 10/14/24 10:46 PM PST. Added schematic to clarify conditions when these readings were taken, and show calculations
-
I look at probing grids like Schrödinger looks at cats in a box
you can find valid data, you can fin erroneous data, or you can cause unintended consequences
a real handy feature on my Fluke is the min/max/avg you can monitor while wailing away for 15 minutes, then kick back and do some easy math, tweak if necessary then play like you stole it
One time, I had a good unintended consequence from probing a grid in a non-working amp--the amp started working. Turned out I had omitted the grid leak resistor.
-
If the choke can cope with about 60mA, I suggest to move swap it to be between the A and B nodes. Try various values 1k to 10k between the B and C nodes to achieve the desired 6V6 conditions.
The choke I used is the Fender Deluxe Reverb style which I believe is 4H 50ma. Otherwise I would make the change you suggest. I'll look into the cost of an upgraded choke.
-
I really appreciate all the responses.
I increased the screen resistor to 9K (added an 8K in series, the biggest 2 watter in my stash). ...
My understanding is that a high value resistor, in series with the screen grid, will cause local negative feedback, akin to a non bypassed cathode resistor.
And a change that makes things a bit quieter is generally perceived as sounding worse.
-
Just adding to Merlin's note on DMMs.
I have a Fluke 179 True RMS DMM. Definitely good for 50Hz, 100Hz and Very useful for chasing tone to 1kHz.
See attached specification. Sluckey used a Fluke 87, even better. :laugh: But both DMMs do cost !!
Basic/Cheaper DMMs Ok at 50/100Hz But I wouldn't rely on them for chasing audio.
Regards
Mirek
-
And.....off down the rabbit hole I go.
So I disconnected the screen resistor(s) and subbed in a 10K dropping resistor in place of the choke. So more similar to 5F1 Champ. The effect was pretty dramatic.
Plate up 345 V
Screen way down 216 V
Cathode way down 13.3 V
Dissipation down to 9.6 W
I then changed cathode resistance from 458R down to 271R, giving me
Plate 334V
Screen 203 V
Cathode 10.1 V
Dissipation 12.3 W
Without the screen resistor, I couldn't measure screen current.
Eliminating the choke did not seem to cause any noticeable increase in hum.
I did not get a chance to play very long after these changes. But the top of the volume range was not as harsh as with the 9K screen resistor configuration.
I'm kind of thinking the 10K dropper is overkill in this application. I'll probably try tacking another 10K in parallel and see what that does.
-
... I disconnected the screen resistor(s) and subbed in a 10K dropping resistor in place of the choke. So more similar to 5F1 Champ. The effect was pretty dramatic.
Plate up 345 V
Screen way down 216 V
Cathode way down 13.3 V
Dissipation down to 9.6 W
As screen grid voltage is reduced, anode current will reduce, hence, given HT supply resistance / sag, anode voltage will increase.
...
Without the screen resistor, I couldn't measure screen current.
Eliminating the choke did not seem to cause any noticeable increase in hum.
Aiken is a strong advocate for 1R cathode current sensing resistors.
1R or 10R current sensing resistors can equally be used elsewhere, eg in series with the screen grid in this case.
The inductive reactance of a 4H choke at 120Hz is about 3k ohms. Hence a 10k resistor here might be expected to provide better smoothing.
Plus lowering the screen grid voltage and anode current will tend to reduce stage gain.
So a culminative effect of less amplification of a lower ripple.
-
The numbers in reply #12 do not add up.
345v-216v is 129v/10k is 129ma.
13.3v/468 is 29ma.
So, I assume 216v is 316v?
If using a 6v6 it is usually recommended to keep the cathode resistor 470ohm (although some increase it to cool down bias a bit).
5e1(choke),5f1(10k)aa764(1k) all work well. If following the schematic in OP, I would ditch the 100ohm resistor between A and B node.
-
The numbers in reply #12 do not add up.
345v-216v is 129v/10k is 129ma.
The decimal got lost: 129v / 10kΩ = 12.9mA
However, most numbers in much of the thread don't add up. I imagine that's "Resistors present but not stated" and/or "resistors not measuring the assumed/stated values."
... 5f1 ... aa764 ... all work well. If following the schematic in OP, I would ditch the 100ohm resistor between A and B node.
5F1 (https://el34world.com/charts/Schematics/files/Fender/Fender_champ_5f1_schem.pdf) and AA764 (https://el34world.com/charts/Schematics/files/Fender/Fender_champ_aa764_schematic.pdf) have more hum than 5E1 (https://el34world.com/charts/Schematics/files/Fender/Fender_champ_5e1.pdf), 5F2 (https://el34world.com/charts/Schematics/files/Fender/Fender_princeton_5f2.pdf), 5E2 (https://el34world.com/charts/Schematics/files/Fender/Fender_princeton_5e2.pdf), and all earlier Champ & Princeton amps. That's because they attach the output transformer to the 1st filter cap after the rectifier rather than at a later stage of filtering.
Personally, "100Ω" doesn't seem a worthwhile resistance, and I would do either 470Ω or a choke.
-
The numbers in reply #12 do not add up.
345v-216v is 129v/10k is 129ma.
The decimal got lost: 129v / 10kΩ = 12.9mA
However, most numbers in much of the thread don't add up. I imagine that's "Resistors present but not stated" and/or "resistors not measuring the assumed/stated values."
... 5f1 ... aa764 ... all work well. If following the schematic in OP, I would ditch the 100ohm resistor between A and B node.
5F1 (https://el34world.com/charts/Schematics/files/Fender/Fender_champ_5f1_schem.pdf) and AA764 (https://el34world.com/charts/Schematics/files/Fender/Fender_champ_aa764_schematic.pdf) have more hum than 5E1 (https://el34world.com/charts/Schematics/files/Fender/Fender_champ_5e1.pdf), 5F2 (https://el34world.com/charts/Schematics/files/Fender/Fender_princeton_5f2.pdf), 5E2 (https://el34world.com/charts/Schematics/files/Fender/Fender_princeton_5e2.pdf), and all earlier Champ & Princeton amps. That's because they attach the output transformer to the 1st filter cap after the rectifier rather than at a later stage of filtering.
Personally, "100Ω" doesn't seem a worthwhile resistance, and I would do either 470Ω or a choke.
Yikes! All the voltages I have posted were measured directly with my Extech DMM in the amp as it currently exists. It is not a Fluke level instrument, admittedly. I will write up a sheet with my calculations and post it tonight. If I have erred, it would be great to know.
As far as the 100 ohm between Node A and Node B, that came from the AX84 schematic I followed originally. I appreciate the suggestion to change to 470 R or a choke.
-
The numbers in reply #12 do not add up.
345v-216v is 129v/10k is 129ma.
13.3v/468 is 29ma.
So, I assume 216v is 316v?
If using a 6v6 it is usually recommended to keep the cathode resistor 470ohm (although some increase it to cool down bias a bit).
5e1(choke),5f1(10k)aa764(1k) all work well. If following the schematic in OP, I would ditch the 100ohm resistor between A and B node.
Thanks for your interest.
345 was the measured plate voltage with the 10k dropping resisor in place.
216 was the measured screen voltage with the 10k dropping resisor in place.
I am embarassed to say I don't know what calculation you are doing by subtracting screen voltage from plate voltage and then dividing by 10k.
-
Reply #7 shows a 9k resistor and 28v drop (327-299).
I checked my amp with a 10k resistor and 41v drop(412-371).
Reply #12 shows a 10k resistor and 129v drop(345-216).
A 10k should not drop that much voltage.
I am speaking generally when I say 5f1 style amp. I have built a few with and without chokes, with just a volume control or with volume and tone or bass , treble, speed intensity,etc. With 16uf/8uf/8uf, or 20uf/20uf/20uf or 20uf+20uf/16uf/16uf or various other values. They all sound good.
I would start with a known good schematic/power supply like 5e1,5e2, 5f2 and possibly slightly increase filter caps.
-
Reply #7 shows a 9k resistor and 28v drop (327-299).
I checked my amp with a 10k resistor and 41v drop(412-371).
Reply #12 shows a 10k resistor and 129v drop(345-216).
A 10k should not drop that much voltage.
I am speaking generally when I say 5f1 style amp. I have built a few with and without chokes, with just a volume control or with volume and tone or bass , treble, speed intensity,etc. With 16uf/8uf/8uf, or 20uf/20uf/20uf or 20uf+20uf/16uf/16uf or various other values. They all sound good.
I would start with a known good schematic/power supply like 5e1,5e2, 5f2 and possibly slightly increase filter caps.
Thank you. I added a document to reply # 7 showing the configurations when those readings were taken. Note that the 9K is a screen resistor, attached to 6V6 pin 4, and the power supply still had the choke between node B and C.
As to reply 10, the choke was removed and replaced with a 10K dropping resistor, and the 6V6 pin 4 was connected directly to Node C. So nodes B and C and the screen connection are the same as Node A and B on the 5F1. So yeah, why such a big voltage drop after those 2 changes? The resistor I used is one of the white recctangular NTE 5 watt (wire wound AFAIK). It measures pretty close to 10 k on the bench, could the resistance change with +/- 300 volts applied? Edit: the resistor is ok, just too large for the application, causing higher than desired voltage drop.
-
... So yeah, why such a big voltage drop after those 2 changes? ...
The 9k is only carrying screen grid current.
Whereas the 10k is carrying everything other than 6V6 anode current.
-
... So yeah, why such a big voltage drop after those 2 changes? ...
The 9k is only carrying screen grid current.
Whereas the 10k is carrying everything other than 6V6 anode current.
Yes. I pulled 2 of the preamp tubes and the voltage and the voltage drop went down. I also tried a different 6V6 and the voltage drop was less.
Also, I tried a different resistor in place of the 10k dropper and the 10k resistor is fine.
-
Am I right in thinking I need to increase the screen resistor substantially?
No. A vintage 6v6 at 340vdc does not need a screen resistor connected to pin4. One can be used if wanted, but a more typical value is 470ohm. Anything over that starts to effect sound quality for a 6v6.
If a 470ohm screen resistor is used put in between pin4 and pin6. Then, either use the choke between b node and c node or use a resistor 1k to 10k between b node and c node.
Or getting a 60ma or greater choke and put it between a node and b node.
-
A vintage 6v6 at 340vdc does not need a screen resistor connected to pin4. One can be used if wanted, but a more typical value is 470ohm. Anything over that starts to effect sound quality for a 6v6.
That is fairly low dcv, we know Fender ran them safely at ~420dcv in Deluxe Reverbs. Way above the 6V6 spec sheets.
But the screen is just a very thin piece of wire, it's not like the plate that has a much larger surface area that dissipates heat. So we want to be careful with the screens max dissipation.
And with new manufacture tubes they don't seem to be as robust as old manufacture tubes.
So, a lot of people always put at least a 470R for the screen. KOC, London Power amps, recomends a 1K screen R. And if you like a little tube compression, going with 1K or a little more, will get you some.
-
So, a lot of people always put at least a 470R for the screen. KOC, London Power amps, recomends a 1K screen R. And if you like a little tube compression, going with 1K or a little more, will get you some.
OK, cool. I'll reconnect the 1K screen resistor with a switch to bypass and see if I can tell the difference.
I saw a youtube recently (so it must be true, right?) that said Fender has a 470R screen resistor in the late model " '57 "Champs.
-
Fender and many other tube amp builders/companies used 470 ohm in most/all of their amps for the screen R, it was and still is the standard.
-
...
Aiken is a strong advocate for 1R cathode current sensing resistors.
1R or 10R current sensing resistors can equally be used elsewhere, eg in series with the screen grid in this case.
Found a spot for a screen current sensing 1 ohm resistor. Didn’t have to drill any new holes
-
Why do you think you need a current sensing R for the power tube screen?
-
Why do you think you need a current sensing R for the power tube screen?
As I fumble & stumble through this project I want to see what screen current is doing. Just out of curiosity.
Plus it was an excuse to make another mini turret board
-
Tried some lower and lower dropping resistor values between Node B and Node C. I had to improvise some parallel combinations, and ended up at 1.75K, which puts dissipation at 12.2W, and puts the screen 31v below the plate. So not too far off from the Champ 5F1.
On the chart I included the numbers from the previous setup with the choke and screen resistor. Not an apples to apples comparison, because 6V6 cathode resistance was different.
-
You can't do the heater elevation like that.
You need to inject the elevation dcv through the CT or a faux CT, 2 x 220 ohm R's.
https://www.valvewizard.co.uk/heater.html (https://www.valvewizard.co.uk/heater.html)
And you should fix your 2 cathode current readings, would clearer if they were labeled mA.
-
You can't do the heater elevation like that.
You need to inject the elevation dcv through the CT or a faux CT, 2 x 220 ohm R's.
https://www.valvewizard.co.uk/heater.html (https://www.valvewizard.co.uk/heater.html)
And you should fix your 2 cathode current readings, would clearer if they were labeled mA.
Thanks Willabe. I will look at the heater elevation info. I guess whoever drew up the original AX84 schematic did not get the memo. I've had it wired that way for many years.
Screen current corrected on diagram.
-
I've had it wired that way for many years.
You can do it that way, but you should get better hum rejection with a CT.
-
Single ended heater circuit wiring will tend induce hum in circuits in close proximity, whereas a balanced circuit won't.
It's a completely different thing than heater circuit DC elevation, which mitigates for the imperfect heater-cathode insulation of some valves.
-
I tried 3 different tubes today. 5V4 rectifier used for all. Node B to Node C dropping resistor is 1750. Rythm channel used for all readings. Volume on 7.5, Master Volume on 10
The "stress" screen current readings are rapid strumming an open A chord on bridge + middle PAF style humbuckers.
-
Greetings from down in the rabbit hole!
I added the 10kEDIT 1K screen resistor back in, but with a bypass switch.
I took measurements with the same three tubes, at 3 different values of cathode resistor. In addition to the traditional 470 ohms as in the Fender Champ, I wanted to try hotter and colder bias.
To me it is interesting to see the similarities and differences in where the various parameters line up. These are all used tubes, I would guess age makes a difference.
In the Champ-iverse, players seem to want lots of power amp distortion, the earlier the better. The Gear Page thread linked in the first post of this thread points out how cathode voltage affects the power amp overdrive point. And specifically how the higher voltage champs end up with higher cathode voltage that doesn’t give the desired distortion character.
For this amp, I’m actually thinking I want the power amp overdrive point as high as possible, within the limitations of a SE 6V6. I want the distortion to come from the preamp. So I think I want higher cathode voltage. From the table it looks like the higher cathode resistor gives m a couple volts, and maybe I can get away with going even higher 6v6 cathode resistor. Although it appears that as the cathode resistor increases, dissipation goes down, so maybe I don’t actually get that much benefit?
Am I completely out in left field here?
-
I want the power amp overdrive point as high as possible,
as you've figured out, "There's no free lunch"
I typically set up the idle tube volts so I have the biggest Signal swing without distortion. Once that's set, I go back to the drive side and adjust so that the drive signal is ~~~~~~~~~~20% Above the PA tubes "clean-swing". Verify, then hand it to guitarists to test drive, including pedals, multiple guitars, speakers .....
listen to their evaluation, then re-weak or call it good.
-
In the Champ-iverse, players seem to want lots of power amp distortion, the earlier the better. The Gear Page thread linked in the first post of this thread points out how cathode voltage affects the power amp overdrive point. ...
For this amp, I’m actually thinking I want the power amp overdrive point as high as possible, within the limitations of a SE 6V6. I want the distortion to come from the preamp. So I think I want higher cathode voltage. ...
Am I completely out in left field here?
Turning the Master Volume down doesn't cause the preamp to distort? It would seem like that's the first thing to try.
Apart from the Master Volume, I only see place-holders for the preamp "schematic" so that doesn't leave a way to evaluate/estimate what signal-level would cause a given preamp stage to distort. Knowing when that occurs ("onset of distortion for last preamp stage is at ___ volts of output") would at least tell you how big a signal the power tube needs to cope with while staying clean.
There's a different issue: You're not drawing a loadline for the output tube and investigating what condition causes it to have its maximum clean output power.
- The plate & screen voltages, and output transformer primary impedance establish parameters you're working within.
- One end of the loadline for the OT primary impedance is at "0v plate-to-cathode" and distortion caused by the onset of grid-current.
- The other end of the loadline for the OT primary impedance is at low- or zero-current for the 6V6, and distortion caused by plate current cutoff.
- Maximum clean output power happens when the 6V6 is "center-biased" and uses the cathode resistor that lands at an idle current that's halfway between each of the extremes.
- Going ever-higher on the cathode resistor value might help at first, but then it starts moving the idle point closer to "plate current cutoff" which will reduce the clean power output.
So you can not-design and try various cathode resistor values. You'll eventually stumble on a value that gives a result you like.
Or draw a loadline for the supply volts & OT primary impedance, and figure out the idle plate current for center-bias, and the required grid-volts to land at that point. From there, Grid-Volts / Idle Current (Plate + Screen) = Cathode Resistor for Center-Bias.
As for the TGP thing: When your only tool is a hammer, every problem looks like "a nail." I felt like they should put "screen volts" in their toolbox for reducing 6V6 plate dissipation, alongside the "cathode resistor value" tool they already had.
-
Thanks HBP. Im trying to work through Merlin's SE tutorial. So far I can see how OT impedence affects things, and thinking about how best to use the available OT taps. and supply voltages.
I guess I'll be in the rabbit hole for a while longer.
-
First time doing a power tube load line. This not the final load line by any means. Just me getting familiar with the process.
I reached my first stumbling block with Merlin's statement
".....the HT in a Class A amp must never be more than half the maximum peak anode voltage rating of the valve, given on the data sheet. For the EL34 this is 2000V so we are well within safe limits!"
My RCA tube manual does not list a "maximum peak anode voltage rating" for a 6V6. For that matter I looked at the 6CA7/EL34 page and it does not list a "maximum peak anode voltage rating" for that tube either. So I don't know where to get this value.
Can anybody help me find the "maximum peak anode voltage rating" for the 6V6?
Thank you
-
I wouldn't worry about it, enough 6V6 amps have been built that we know what the practical limitis are. As long as the B+ is below 400V you won't have a problem. I should really update that page. If you want maximum clean power the best option is generally to copy the datasheet recommendations.
-
This datasheet says peak pulse 1,200V (which is not strictly for audio application, but it’s a number that may satisfy your curiosity)
https://frank.pocnet.net/sheets/135/6/6V6GTA.pdf (https://frank.pocnet.net/sheets/135/6/6V6GTA.pdf)
-
First time doing a power tube load line. This not the final load line by any means.
Your chart is for a screen voltage of 285. If you want to run the screens closer to the plate voltage in your amp, you’ll need to plot an appropriate set of grid curves. (Which will be higher up.) This will end up with the load line crossing well below the knee of Vg0 which will more likely to encourage high screen current under big signal conditions, and why a screen resistor might be a good idea.) Keeping Vg2 lower is a way of keeping the screens happier.
-
Keeping Vg2 lower is a way of keeping the screens happier.
but not always the guitarist, the audiophile, ya, those rockers though a finicky bunch :icon_biggrin:
-
Thanks Tubeswell. I noticed about the 285 screen voltage on the chart. The only other blank 6v6 charts I found have even lower screen voltages. If you have a chart appropriate forI searched for higher screen voltages could you post it up?
I think I understand how to use the B+ and transformer pimary impedence to get the initial lines on the chart, and select an operating point below the safe operating limit curve. But I am struggling to grasp the the next steps having to do with selecting screen voltage. I have learned how to adjust screen voltage via the Node B to Node C dropping resistor, but I dont know how to select the best screen voltage.
The RCA "Typical operating conditions" seem to indicate that for plate voltage of 250v, it is ok to have screens at the same voltage, but at 335v that it is advantageous to have the screens all the way down to 220.As it happens my screen voltage is sitting at 282v with the 5Y3 rectifier, pretty close to 285. With plate at 309V, cathode 17.8V.
One more question, what does keeping the load line below the knee mean? I looked up the definition of "knee of a curve", and the load lines I drew do not seem very close to the knee of the 0V grid curve. Thanks again
-
These screen curves came straight off my curve tracer, they show the 0V grid curve for different screen voltages. These are handy for finding the knee (of the 0V grid curve, as your correctly observed)
-
Keeping Vg2 lower is a way of keeping the screens happier.
but not always the guitarist, the audiophile, ya, those rockers though a finicky bunch :icon_biggrin:
Yup - ‘s why I carefully chose my words :-)
@pullshocks - the knee is the bendy bit of the Vg0 grid curve. FWIW, if you don’t have Merlin’s curve plotter, you can manually plot ‘good enough’ grid curves for whatever g2 voltage you are ever likely to need extrapolating from the g2 voltage curves on average transfer characteristics chart for plate current (e.g., top of P6 in the datasheet linked above).
-
Very cool Merlin. Much appreciated.
I tried to sketch in a 0 grid curve for 285v screens
And *roughly* copied my line for 8K impedence and 310v plate voltage. So if I understand this, it would be OK for screen voltage down as far as 250v but not below?
-
Keeping Vg2 lower is a way of keeping the screens happier.
@pullshocks - the knee is the bendy bit of the Vg0 grid curve. FWIW, if you don’t have Merlin’s curve plotter, you can manually plot ‘good enough’ grid curves for whatever g2 voltage you are ever likely to need extrapolating from the g2 voltage curves on average transfer characteristics chart for plate current (e.g., top of P6 in the datasheet linked above).
Tubeswell, I only have the RCA tube manual, not the full data sheet. It is only a page, no graphs or charts. Do you have a copy of the full data sheet?
-
@pullshocks - the knee is the bendy bit of the Vg0 grid curve. FWIW, if you don’t have Merlin’s curve plotter, you can manually plot ‘good enough’ grid curves for whatever g2 voltage you are ever likely to need extrapolating from the g2 voltage curves on average transfer characteristics chart for plate current (e.g., top of P6 in the datasheet linked above).
Tubeswell, I only have the RCA tube manual, not the full data sheet. It is only a page, no graphs or charts. Do you have a copy of the full data sheet?
He posted the link in reply 41.
https://frank.pocnet.net/sheets/135/6/6V6GTA.pdf
-
And here's some on line copies of some tube manuals;
RCA - Hallman Labs (https://hallmanlabs.com/catalogs/rcatubes/)
https://hallmanlabs.com/catalogs/sylvania/ (https://hallmanlabs.com/catalogs/sylvania/)
https://hallmanlabs.com/catalogs/general-electric/ (https://hallmanlabs.com/catalogs/general-electric/)
https://hallmanlabs.com/wp-content/uploads/2020/01/ETR-15G_General_Electric_Essential_Characteristics.pdf (https://hallmanlabs.com/wp-content/uploads/2020/01/ETR-15G_General_Electric_Essential_Characteristics.pdf)
-
Also RCA manuals say look at 6AQ5 tube for chart. Their related.
-
Also RCA manuals say look at 6AQ5 tube for chart. Their related.
Thanks a lot for all those links, and to Tubeswell, I did miss seeing the link in post 41.
You are right, RCA manual does reference the 6AQ5 page, but the graphs on that page are pretty minimal, no substitute for the full meal deal data sheets
-
And *roughly* copied my line for 8K impedence and 310v plate voltage. So if I understand this, it would be OK for screen voltage down as far as 250v but not below?
It would be fine for a screen voltage down to 200V but below that you would lose power. You see Champ-alikes with screen voltages up to 300V with a 8k load, so I wouldn't sweat it too much
-
I wonder why it came about that 6L6x info is so extensive (numerous different charts, typical operating conditions), whereas 6V6x info is pretty 'bare bones'?
Online loadline plotters / calculators are handy for 6V6, if one's anode characteristics chart extrapolation skills are somewhat lacking :)
https://www.vtadiy.com/loadline-calculators/loadline-calculator/
http://bmamps.com/ivds.html
-
I wonder why it came about that 6L6x info is so extensive (numerous different charts, typical operating conditions), whereas 6V6x info is pretty 'bare bones'?
6L6x bigger/more power, better, later developed tube.
Trying to sell the newer design 6L6 series, and leave the older design 6V6 behind.
-
..... if one's anode characteristics chart extrapolation skills are somewhat lacking :)
Thanks, I really appreciate it.
I've tried, but I'm still "somewhat lacking." The on line calculators are helping, but I'm confused about a couple aspects:
BNAmps--
1) calls for Plate to cathode voltage --It seems like this field should be plate voltage. You start with a selected plate voltage, but the EDIT plate to cathode voltage depends on cathode voltage, which depends on cathode resistor value, so how do you know in advance what the plate to cathode will be?
2) The notes at the top state "a) The bias control only applies to the push-pull option" but I found that changing the value in the grod bias field does significatly affect the end result, with -20v being the most favorable for 6V6 SE transformer load in the cases I tried.
VTADIY--
What info goes in the field "Out.headroom (+/-V)%"
https://www.vtadiy.com/loadline-calculators/loadline-calculator/ (https://www.vtadiy.com/loadline-calculators/loadline-calculator/)
http://bmamps.com/ivds.html (http://bmamps.com/ivds.html)
Thanks again.
-
but the cathode voltage depends on cathode voltage, which depends on cathode resistor value, so how do you know in advance what the plate to cathode will be?
and now you "see" the difference between design, and reality.
in "design" mode, you start with the OT primary impedance and the "target" plate VDC. Do the math and you have "Theoretical" tube current. now with the cathode R, the tube current, more math, you're "in the ballpark", turn it over to the lab tech, have it breadboarded up, then measure and adjust
-
how do you know in advance what the plate to cathode will be?
You lick your finger and stick it in the air. 6V6 bias under 'typical' conditions is in the range of 15 to 20V, so you just subtract that from the B+ and call that your plate-to-cathode voltage. It'll be close enough.
-
[, turn it over to the lab tech, have it breadboarded up, then measure and adjust
Yes things are in the hands of the ...uh... lab tech. As best as he can tell with his Extech DMM, it is coming out within 20% of predicted values. He is working up a budget request for an oscilloscope and signal generator
-
...
VTADIY--
What info goes in the field "Out.headroom (+/-V)%"
...
Just leave it blank.
-
how do you know in advance what the plate to cathode will be?
You lick your finger and stick it in the air. 6V6 bias under 'typical' conditions is in the range of 15 to 20V, so you just subtract that from the B+ and call that your plate-to-cathode voltage. It'll be close enough.
Thank you Merlin. That is what I surmised. Also, it only accepts plate and screen volts in 5 volt increments
But trying different combinations of inputs has helped me visualize the interplay.
-
...
VTADIY--
What info goes in the field "Out.headroom (+/-V)%"
...
Just leave it blank.
I tried entering some numbers. Once you enter a number, it displays a power output. With plate voltage of 330, Entering the number 290 in the out headroom box gives a power output of 5.25W and THD of about 12, which lines up with the RCA Tube Manual typpical operating conditions.
-
My 2 cents taken from building single ended amps: If the voltages and and currents are close to the datasheet everything it is fine, no big differences to the sound when having slightly different values there. But in one build i had only the screen supply of a SE output on a separate power-transformer winding and this gave a unique sound. I would not want this sound in an output tube but for a reverb driver it is very cool, it has very little sag, i would describe it, like walking shoes that go through any obstacle. I know SE output stages "don't sag" but it really made a difference to me. I do keep track of all my voltages while designing with soldering iron but little do i know about EE, so add your grain of salt please.
-
..... But in one build i had only the screen supply of a SE output on a separate power-transformer winding and this gave a unique sound.
Ok, but fill in the blanks please.
What was the SE power tube, what plate dcv, what was the screen dcv, K R, and how many watts output with the set up?
-
My 2 cents taken from building single ended amps:
6V6, 6K6, EL84, 6L6, EL34, 6550, KT88 xSE builds, the most "versatile" SE amp has a Power amp with a stiff SS PS, an oversized OT with 40% taps into the best speaker money can buy.
This gets you Class A PA, drive it with whatever "front end" you choose, keep the drive signal "available" at ~~~ 20% over PA tube datasheet for "max signal G1"
you can add all the flavor you want, up front, an SE PA will just pass it to the speaker unmolested by PA builders creativity.
If you just have to OD the PA for "that sound" the 20% extra drive signal is there while keeping the PA "in-check", more than that and 8 outta 10 times the PA heads for some death metal sounds that are real hard to "design out"
-
Thank you Shooter and Williamblake. Here’s my 2 cents, to go with your 2 cents.
After studying Merlin’s SE page and playing with the simulators, I decided I should change the output transformer to 8K. It had been on a 5K tap. It took a while to make that change, so I couldn’t do much of an A/B comparison, but I think it sounds better. Or it could be confirmation bias rearing its ugly head again.
I also think it sounds better with the 5V4, but of course that also affects voltages in the preamp and cap drive reverb, so not a controlled experiment. Either way, it is set up using a cathode resistor close to the BMAmps calculated value for my voltages and OT impedence. It is not overdissipating. It sounds good. I get good feelings as I play. I’ll probably try tweaking RK, to see if I can hear a difference, but I doubt I’ll be able to tell.
However, beyond certain settings of preamp volume and master volume, I don’t like the sound any more, and I think this is what Shooter is getting at. I can’t find a spec for “max signal G1” on the datasheet, and don’t know how/don’t have the equipment to measure drive signal anyway. I would not be surprised if my “by ear” setting lines up with Shooter’s Law of Single Ended Amps
Speaking of signal level, in the BMAmps simulator, there is a data field for “Signal Input” that defaults to 66.6 volts. For quite a while I just left that value, but then started playing with it and eventually worked up the graph below.
Attached is my latest block diagram, showing revised power supply dropping string and updated voltages, bias resistance, and OT connection .
-
I can’t find a spec for “max signal G1”
try Peak signal G1 the 6v6 sheet I looked at had it at ~~13vac at plate 315 NOTE: peak, is only 1/2 the sine wave, so when you use a scope n sine wave you're looking at ~~26 vac p-p sine wave.
you really need to sell a kidney or something n buy a scope if you want to keep tinkering at this :icon_biggrin:
-
you really need to sell a kidney or something n buy a scope if you want to keep tinkering at this :icon_biggrin:
You've been very generous with your help, so I hate to ask another favor, but, OK, help me pick (or suggest something else adequate for this work, and easy to learn to use)
Leader 8103 100MHz Oscilloscope + Ham Radio CB RF Sampler - electronics - by owner - sale - craigslist (https://seattle.craigslist.org/est/ele/d/seattle-leader-mhz-oscilloscope-ham/7792077672.html)
Hantek DSO2C10 Digital Storage Oscilloscope 100MHz Bandwidth 2CH Dual Channel 1GSa/s 8M Memory Depth with High Voltage Probre: Amazon.com: Industrial & Scientific (https://www.amazon.com/Hantek-DSO2C10-Digital-Oscilloscope-Bandwidth/dp/B08Y6SKTX6/ref=sr_1_2_sspa?crid=6M9YBWZ1NT6N&dib=eyJ2IjoiMSJ9.o8lL8s5TUSQtT78oJ3hLvnP3L_XhorMvfyzkDRa7qCnxi59wETBfH-yZSU6zxQBJ4TTdEluXWGtKdauQ2b0aFEdqfQWNIiZFXiJL4nLFz3ombMjHyiBg1Z0Nx_DbX1YPQ308iJ7BJBRqu5Ei2Fu1nlOFGC6hS_NUC23HMZRSk4cpGMurkPSx8jGYO_mXz9_fhbs3Y9Yn3Ata6b26tdkum-7ABH5uhKKMbYXK8Cv11O4.1rqJXgGoZhb70vGBTnRUdEMhB_HhW01hhmGAfp9eCnM&dib_tag=se&keywords=oscilloscope&qid=1730244484&refinements=p_36%3A7000-22000&rnid=2661611011&sprefix=os%2Caps%2C502&sr=8-2-spons&sp_csd=d2lkZ2V0TmFtZT1zcF9hdGY&th=1)
Handheld Pocket Digital 120 MHz Color Oscilloscope Replaces Velleman HPS140MK2 (https://vetco.net/products/handheld-pocket-120-mhz-digital-color-oscilloscope-single-channel)
-
Not the handheld one. I'd probably lean towards the analog scope for tube amp work, since you get a 'smoother' image. But digital is great for giving you instant readouts of voltage, frequency, true RMS etc. Real gangsters have one of each.
-
what's your long game?
will you be doing repairs for others?
do you tinker in the digital realm, breadboarding, micro-controllers....?
simply to "see" what's up on your builds?
unless the scope is real cheap AND works fine, I would avoid CRT based scopes like the 1st pic, you can tell in the image "somethings up" by the uneven trace intensity, normal on left, then bright.
the 2nd one looks at 1st glance to be a good all-round "bench scope" something you'll use on a regular basis for repairs, analysis, design.
the last one is closest to "just a good tool to have when I need it"
this one was "reviewed" by Sluckey awhile back, and the price is good for that quick look at things, quick and handy for tweaking but I wouldn't put it in the "I use a scope often" group.
https://el34world.com/Forum/index.php?topic=22702.0 (https://el34world.com/Forum/index.php?topic=22702.0)
-
you can tell in the image "somethings up" by the uneven trace intensity, normal on left, then bright.
That's just a camera artefact. I don't live in America, but $200 sound like a lot for a secondhand scope like that. OK it's 100MHz, but it looks well-used and otherwise bog-standard. $100?
-
what's your long game?
Thats the hell of it, this is not a long game proposition. I want to dial in this project and then get out of the tube amp building hobby. So buy, use it for a month or 2, sell for whatever I can get back out of it, no doubt at a loss. After looking into this, I also need a signal generator, a bucket of probes, cables, and adapters, and apparently an isolation transformer
Starting to think it would be more cost effective to take it to a tech, if I can find one who is willing to work with me
-
this is not a long game proposition.
Starting to think it would be more cost effective to take it to a tech, if I can find one who is willing to work with me
we're free here, otherwise I charge $100 over cost for the no frills, add trem add $80 add verb....
post your design idea schematic
build it
debug the VDC and "gross" signal issues
match up a speaker
play for 1 month
take it to a tech for "tweaking" and dialing it in.