Hoffman Amplifiers Tube Amplifier Forum
Amp Stuff => Tube Amp Building - Tweaks - Repairs => Topic started by: AmberB on February 24, 2022, 01:08:21 am
-
Another question that I'm not sure about.
If you use a 1 ohm resistor on the cathode of the power tubes in an amp with a negative bias, what wattage rating should that resistor have? On a 6L6 tube, for instance?
-
Normally the current is not that big so 1W or even lower is enough. But if something goes wrong with the power tubes the cathode is the first thing that blows. That's why I recommend 5W resistors.
/Leevi
-
1/2 watts resistor may protect amp like a fuse.
I always put 1/2 watt
Cathod current is often less than 50 ma
On a 1 ohms resistor current and voltage are the same
50ma X 50 mv = 0.0025 watts
A 0.5 watts resistor is big enough
-
+ 1 for Latole
Franco
-
Normally the current is not that big so 1W or even lower is enough. But if something goes wrong with the power tubes the cathode is the first thing that blows. That's why I recommend 5W resistors.
/Leevi
To blow a 1/2W 1ohm resistor, more than 700mA of HT current has to flow. That’s a lot!
If the resistor doesn’t blow, consider that the next thing to get damaged by such fault current will almost inevitably be more expensive / difficult to fix.
See https://www.thegearpage.net/board/index.php?threads/fender-64-custom-deluxe-reverb-handwired-schematic.1918269/page-2#post-26115004
My view is that best practice is to fit a diode across the 1ohm resistor, in conjunction with ensuring that the HT will fuse quickly when fault current is drawn from it.
Some modern Fenders use a 1W 1ohm resistor with parallel diode, eg https://el34world.com/charts/Schematics/files/Fender/Fender_65_super_reverb.pdf
-
I had the same question not too long ago...here's a link to that thread with a little more info:
https://el34world.com/Forum/index.php?topic=28427.msg312935#msg312935 (https://el34world.com/Forum/index.php?topic=28427.msg312935#msg312935)
-
1/2 watt 1 ohm 1% is all you need
Calculate the power dissipation through a ohm resistor to see why.
-
Historically, 1/2 and 1/4 parts only went to 10 Ohms, not 1 Ohm. That's why we expect to have to order a 1 Watt part.
A little math will show that even 1/10th Watt is unlikely at tube currents.
-
Some modern Fenders use a 1W 1ohm resistor with parallel diode,
am I correct assuming the 1ohm needs to blow in order for the diode to blow?
-
Some modern Fenders use a 1W 1ohm resistor with parallel diode,
am I correct assuming the 1ohm needs to blow in order for the diode to blow?
The diode should stop the resistor from ever blowing. ie when the current exceeds 600-700mA, the diode will conduct. So the resistor shouldn’t dissipate much more than 1/2W not matter what the fault current.
When diodes fail, they tend to fail short circuit, rather than open.
-
Thank you for all the replies!
-
10 ohm may be a better technical choice, and is what I prefer to use.
10x more power dissipated in the part means it is more likely to act as a poor-mans fuse. 10 ohm resistors are readily available in 0.25W rating, making them more likely to act as a poor-mans fuse or a smoke and odour inducing indicator that something is seriously wrong in your amp.
A 10 ohm resistor makes it much easier to use any type of DMM to make a reasonably accurate measurement of cathode current, as not all DMMs can make accurate measurements and resolve 1mA difference down at the 30-50mV level.
-
+1 for EL34 and PRR (and Latole).
We are trying to get a fairly accurate idea of current therefore 1% Resistor accuracy is important (at least it is to me).
My idea is to get the cheapest (and most stable) resistor 1R, 1% in a reasonable compact package.
I find modern 1R, 1% and either 1/2W or 1W Metal Oxide or Metal Film fit the bill.
Point taken from trobbins. I do use a Fluke 79 (and 867B), so I am comfortable with 1R and what you see on display is what you get (simple maths)
Regards
Mirek
-
when the current exceeds
Thanks, I was imagining voltage as the diode "switch" I couldn't get it to .6vdc before the current killed the R
-
i usually use 1w because they're generally physically larger and easier to get my meter's probe (or a clip lead) attached to thicker leads. if you've got external bias measurement points then size isn't as much of a concern so use whatever fits
-
A 10 ohm resistor makes it much easier to use any type of DMM to make a reasonably accurate measurement of cathode current, as not all DMMs can make accurate measurements and resolve 1mA difference down at the 30-50mV level.
That has completely sense to me (also if I've a set of affidable DMMs and a stock of 1R 3W 1% resistors)
Franco
-
when the current exceeds
Thanks, I was imagining voltage as the diode "switch" I couldn't get it to .6vdc before the current killed the R
Voltage IS the switch that causes the diode to conduct.
Say the diode is forward-biased at 0.7v. That happens at 0.7A through the 1Ω resistor.
0.7v x 0.7A = 0.49 watts ---->. Fender's 1w 1Ω resistor is unconditionally safe.
Rather than protect the resistor, I'd prefer to see a circuit that actually protected the tube or output/power transformer.
-
Rather than protect the resistor, I'd prefer to see a circuit that actually protected the tube or output/power transformer.
Bingo! By the time the diode conducts, the tube is probably already shot and the OT is following close behind.
-
Say the diode is forward-biased at 0.7v.
that is my head scratcher, IF normal is say 40mA, you have 40mV across the 1R, no current through the diode.
when the circuit starts drawing LOTS of current, V will drop not rise, the diode still won't conduct, until the R opens, then I goes through the diode, shorts it from excess current, then opens from the heat.
-
that is my head scratcher, IF normal is say 40mA, you have 40mV across the 1R, no current through the diode.
when the circuit starts drawing LOTS of current, V will drop not rise...
Clear your head and rethink. As current increases across a resistor, voltage also increases. E = IR.
Don't over think. :icon_biggrin:
-
that is my head scratcher, IF normal is say 40mA, you have 40mV across the 1R, no current through the diode.
when the circuit starts drawing LOTS of current, V will drop not rise, the diode still won't conduct ...
I believe you're pointing out it will "flicker" as the diode conducting knocks the diode voltage to 0v. The when the diode stops conducting, the current goes through the 1Ω until developing forward-bias again.
Unless there's some special mechanism happening there that helps the Mains Fuse blow, I think we're looking for reasoning to what is already know to be ... questionable.
-
Don't over think.
:laugh:
that's the problem, I thought
I just thought wrongly!
thanks for clearing my brain so 6:20 will be more enjoyable :icon_biggrin:
EDIT: showing my work, hoping for partial credit :help:
-
...
Rather than protect the resistor, I'd prefer to see a circuit that actually protected the tube or output/power transformer.
A HT fuse (perhaps Marshall's greatest contribution the valve guitar amp) can be handy for that :icon_biggrin:
To best protect those things, it should probably be a fast acting type, so typically would need to be after the reservoir cap.
But the way I see it, it's not so much about protecting the resistor as making the amp more resiliant to expected failures, less time on the bench if a valve shorts etc. Replace fuse and valve, check bias, and back into action.
-
... showing my work, hoping for partial credit :help:
You wrote, "1Ω 1/2w."
However, pdf64's example used a 1w resistor:
Some modern Fenders use a 1W 1ohm resistor with parallel diode, eg https://el34world.com/charts/Schematics/files/Fender/Fender_65_super_reverb.pdf (https://el34world.com/charts/Schematics/files/Fender/Fender_65_super_reverb.pdf)
-
If you imagine the tube voltage in terms of ‘height above ground’, represented by a stack of 500-sheet reems of photocopy paper, and you have a fixed bias 6L6 with a plate voltage of 460 biased to 30mA, then the cathode voltage across the 1R resistor amounts to 1 sheet of paper, whereas the plate voltage is 30 x 500-sheet reems of paper stacked together one reem at a time. (Or, if you run your 6L6 at 60mA, 15 reems of paper in the stack). The cathode is pretty much still at ground potential with a 1R resistor, and the resistor barely dissipates any power at all.
-
... showing my work, hoping for partial credit :help:
You wrote, "1Ω 1/2w."
However, pdf64's example used a 1w resistor:
Some modern Fenders use a 1W 1ohm resistor with parallel diode, eg https://el34world.com/charts/Schematics/files/Fender/Fender_65_super_reverb.pdf (https://el34world.com/charts/Schematics/files/Fender/Fender_65_super_reverb.pdf)
-
thanks all, I know where I went wrong, I was just attempting humor :BangHead:
shooter -1 forum +3
-
Rather than protect the resistor, I'd prefer to see a circuit that actually protected the tube or output/power transformer.
Give a look to the Fender Bassman 300W Power Section schematic, the part of the schematic (with fuses) that is above the Notes
(https://i.imgur.com/Ml0bCES.png)
(https://i.imgur.com/wAuxAhd.png)
and, if I remember correctly, there were Marshall amps (fixed bias) with a fuse between cathode and ground
OK, I remembered correctly
(https://i.imgur.com/a8l8zLK.gif)
Franco
-
It looks like Fender expects shorts at the screen grid to be a lot more likely than at the anode.
I don’t think Marshall’s cathode fusing arrangement lasted very long.
My experience / hypothesis is that when a valve shorts and a cathode fuse / resistor blows, and the heater circuit is also DC referenced to 0V common, then when the cathode voltage goes high due to the open circuit, the heater cathode insulation breaks down and the fault current makes its way to (from) 0V common via the heaters.
Hence my view is that it’s best / more resilient for the cathode to retain its connection (eg via resistor//diode) to 0V common.
I think Fender specified a 1W 1ohm cathode resistor because the max forward voltage on a typical silicon rectifier type diode is around 1V. So under sustained fault current, a 1/2W resistor would overheat.
That’s possibly a belt and braces thing, hopefully a fuse (especially a fast acting HT fuse :icon_biggrin:) would prevent prolonged exposure.
-
I think Fender specified a 1W 1ohm cathode resistor because the max forward voltage on a typical silicon rectifier type diode is around 1V. So under sustained fault current, a 1/2W resistor would overheat.
May be all is as you say (I'm not so skilled, only, sometime, I remember how factories do it)
I remembered also that here they put 3 in series diodes instead of only one (and I've no idea for the reason)
I noted that also this is a Big Bass Amp
(https://i.imgur.com/5iHpKLH.jpg)
(https://i.imgur.com/rSLoHt1.jpg)
Franco
-
I think Fender specified a 1W 1ohm cathode resistor because the max forward voltage on a typical silicon rectifier type diode is around 1V. So under sustained fault current, a 1/2W resistor would overheat.
May be all is as you say (I'm not so skilled, only, sometime, I remember how factories do it)
I remembered also that here they put 3 in series diodes instead of only one (and I've no idea for the reason)
I noted that also this is a Big Bass Amp
(https://i.imgur.com/5iHpKLH.jpg)
(https://i.imgur.com/rSLoHt1.jpg)
Franco
Running 3 x 6550 on each side, each set of three all through the same 1R resistor
-
So a 1R 1W resistor on each of two 6V6 is extreme overkill.. and no diode needed
-
So a 1R 1W resistor on each of two 6V6 is extreme overkill.. and no diode needed
Could you explain your thinking on those 2 points?
-
PDF are you following all the posts to this thread?
"Running 3 x 6550 on each side, each set of three all through the same 1R resistor"
The way I look at it is that if something catastrophic happens in a tube something else is going to give before that 1 watt 1 ohm resistor blows.
-
I though I'd share a photo of one of my most valued?/interesting EL34's, removed from a Marshall and I can't remember the model (old age :laugh:)
It obviously got a bit warm when red-plated and the vacuum sucked the glass in :l2:.
Unfortunately it has degassed in the last 5 years with a fine crack appearing around the metalwork.
Obviously the HT fuse didn't go pop and I wonder if a 1R (1/2?) in the cathode pin would have helped?
(A refurbish of the Bias supply and a pair of new tubes and I didn't see the amp again.)
Kind regards
Mirek
-
PDF yes we could run the diode in parallel with the 1 ohm resistor. But lets look at Fender amps from the 50's 60's 70's 80's how many times have you seen one of these amps have a failure as you describe?
I can tell you my experience. Never have I seen that kind of failure. I will run the 1 ohm resistors on the cathodes of my power tubes without even giving thought to your concern. Why? Because I've done it for years with Zero failures as you describe. So although your analysis has merit.... in the real world there is no need for the extra caution because the failure rate for what you describe is next to nothing.
If we built cars for every possible catastrophe that may occur the cars would weigh tons more with 17 air bags per square inch and multiple bumpers ... along with throttle resirictions.. and we would have to wear helmets and bullet proof vests..
There comes a point when we have to look at the costs versus the reality .. how many amps have you seen or experience that have had the failure you describe?
Just had to say it..
-
I think the thread's topic is a bit more wide ranging than Fender amp failures.
I've personally come across at least 10 vintage amps where the OT has been replaced, or had failed, or I bought with no OT in the amp. The OT is plausibly more fragile with respect to output stage valve faults than the PT. Definitely more broad-brush in amp range than one manufacturer.
So I do whatever I can to practically bullet-proof an amp, whether that is to suppress stress causes for starters, or include failure protection such as poor-mans fusing, and that includes adding parts to avoid further collateral damage if a cathode fuse/resistor open-circuits. I am in the camp where I'd prefer to use a 10R cathode resistor of 0.25-0.5W rating, and hope that it opened or told me that something was wrong, rather than alternately designing a 1R resistor to survive by upping its power rating or adding a protective part.
-
Here is a tidbit of history. While cathode resistors for broadcast metering were known long ago, this is one of the earliest domestic-gear schemes.
"1.56V" is reputed to be the voltage of a very fresh C-cell. This actually varies with details of the chemistry, but was often more reliable than 1950s meters.
-
Ciao PRR
Many thanks for posting
that is very interesting
--
BTW
this morning may be I'm steel sleeping, I don't remember what means IM
Franco
-
IM = intermodulation
-
Many Thanks Steve
Franco
-
Thanks PRR. You're a mine of historical knowledge.
-
...they put 3 in series diodes instead of only one (and I've no idea for the reason) ...
Me neither :w2:
On the face of it, that allows up to 3V across the 1ohm 1W resistor under fault conditions, ie 9W! So the diodes seem kinda pointless :dontknow:
PDF are you following all the posts to this thread?...
I thought so; could you point out if you think I've missed something?
...The way I look at it is that if something catastrophic happens in a tube something else is going to give before that 1 watt 1 ohm resistor blows...
Indeed, but then output valves do occasionally fail catastrophically.
My view is that it's good to attempt to limit the collaterol damage arising from likely incidents. So the user just replaces the problem valve, blown fuse, checks bias and the amp is back in use, no need for time on the bench.
PDF yes we could run the diode in parallel with the 1 ohm resistor. But lets look at Fender amps from the 50's 60's 70's 80's how many times have you seen one of these amps have a failure as you describe?
I can tell you my experience. Never have I seen that kind of failure... how many amps have you seen or experience that have had the failure you describe? ...
But those amps didn't generally have 1ohm cathode current sensing resistors :w2:
I suppose we all have to be guided by our personal experience. I've not repaired many amps, valve amp tinkering /fixing etc is just a hobby, so a few dozen maybe over the years. And only a couple of 60s Fenders, being as they're fairly rare here.
But when I'm given something to fix, it plays on my mind how the equipment's design might be amended to reduce the likelihood / consequence of recurrance. Of course valves are always going to short, but it's feasible to make simple changes to the reduce collaterol damage.
I've described a few failure types in this thread, but I'll assume you're referring to the heater - cathode insulation damage to other valves in the amp when an output valve shorts and the heater circuit is not solidly DC referenced to 0V common(?)
I think that's not unusual, at least in my limited experience. Bear in mind that unless the breakdon is total, it might not cause problems in many applications, eg trem oscillator, later stages, earlier stages whose cathodes are fully bypassed. But if used in earlier stages with partially or unbypassed cathodes, the damaged valves cause a buzz to contaminate the signal path.
If we built cars for every possible catastrophe that may occur the cars would weigh tons more with 17 air bags per square inch and multiple bumpers ... along with throttle resirictions.. and we would have to wear helmets and bullet proof vests..
There comes a point when we have to look at the costs versus the reality ..
The fitting of a diode for each cathode current sensing resistor, and a suitably rated, fast acting HT fuse, is all I've suggested; which doesn't seem an inordinate degree of effort to go to?
...
So I do whatever I can to practically bullet-proof an amp, whether that is to suppress stress causes for starters, or include failure protection such as poor-mans fusing, and that includes adding parts to avoid further collateral damage if a cathode fuse/resistor open-circuits. I am in the camp where I'd prefer to use a 10R cathode resistor of 0.25-0.5W rating, and hope that it opened or told me that something was wrong, rather than alternately designing a 1R resistor to survive by upping its power rating or adding a protective part.
It seems fairly common for output valve shorts to result in a heater circuit humdinger or balancing resistors blowing. So unless the heater winding has a CT that's connected to 0V common, on balance I think it's best to try and avoid an output valve short resulting in the cathode losing its connection to 0V common.
Here is a tidbit of history...
Thanks, that reminded me of the mixed bias arrangement used in the later Vox AC50, expect that 47ohm, rather than 12ohm, resistors were used. But then the goal was 47mA idle cathode current, rather than 130mA! :huh: