Supro 1690T build

[dbishopbliss]

I attached the primaries of the PT to a power cable and then soldered the OT secondaries to a jack so I could easily plug in a set of headphones. I plugged the power cable into a power strip with a switch so I could turn it on / off remotely.
 
The photo where the OT is lined up straight is the original location. Hummmmmmmmm. Not terrible but clearly audible.

The photo where the OT is at an angle is almost silent. I had my daughter with "dog hearing" listen for me to find the most quiet position. Moving the OT further away did not make any improvement. Time to drill another hole.

I have to admit, I was surprised by the results. I would have thought that after all these years of people building a JTM45 style amplifier that someone would have figured out that the amp is quieter with an angled OT. Or, is this just a function of this PT/OT combination?

[PRR]

Most musicians don't have a keen-ear daughter on headphones.

[dbishopbliss]

Most musicians don't have a keen-ear daughter on headphones.

At least one that is willing to come into the workshop and listen to buzzzzzz. Full disclosure... I had to drag her to come listen and there was some eye rolling. 

[Willabe]

 

[PRR]

There really is such a thing as a too-strong magnifying glass.

[dbishopbliss]

There really is such a thing as a too-strong magnifying glass.

Perhaps, but even my 55 year old ears could hear the difference.


Different topic, but sort of related. I downloaded the latest version of PSUD2. I am trying to model the power supply for the Supro 1690T so I can compare the differences between using the 20uF/10uF/10uF filter caps as designed and the 32uF/32uF/16uF filter caps that I will be using. However, I keep getting warnings saying, "A current sink has pulled the voltage below zero for more than 5 mains cycles at time..." I think either I am entering some values incorrectly or I am going to need to change the power supply.


Here is what I used:


Primary
Supply Voltage: 125.6 V (measured across the wall socket)
Winding Resistance: 1.7 Ohm (measured primaries end to end)


Secondary
Offload Voltage: 649 V (measured secondaries end to end, not the center tap)
Winding Resistance: 44.5 Ohm (measured end to end, not the center tap)


I read that you should halve the values for calculations when you have a CT. This calculated:
Turns Ratio:  1:2.58
Estimated Impedance: 33.59 Ohms


For the model I used a Full-wave rectifier followed by a C filter, RC filter RC filter, Constant Current Load (seen screen show below):
The datasheet for 6L6 tubes running push-pull class AB1 gave a value of 88mA and each 12AX7 draws 1.5mA for a total of 91.5mA (do I have that right?).


Using all these values, the simulation predicts the voltage at C2 would be around 213V and C3 would be like 18V. In a message from Leon C, the target for the plate of the 6L6 tubes is 362V. So, I think I'm off with something.


Are the values I plugging in to PSUD2 wrong or do I need to adjust the R1 and R2 values. Thanks for your help.


P.S. The screen shot says 91.5A, but it should say mA. Same thing happens.

[shooter]

guessing the 6L6 gets it's "feed" from the C1/R1junction?
if so, that's where you want the high current, not at the end of the PS string

[dbishopbliss]

guessing the 6L6 gets it's "feed" from the C1/R1junction?
if so, that's where you want the high current, not at the end of the PS string

I tried that… I still get the same results.

[shooter]

 
my co-worker Paul loved using modeling software, me I just kept changing things til I found the broke one, I'd be 1/2way home n paul would call, "where'd you go?,  I think it's......"

[dbishopbliss]

I played with the values of the resistors and the following model no longer raises errors and the voltages are in the ballpark of what I think they should be.
R1 has a value of 560 Ohms - WAY down from 10K.
R2 has a value of 20K - double the 10K value.
I have also changed the rectifier to a 5AR4 - I have a couple of NOS Mullard 5AR4 tubes so why not use them?

Again, Leon C said the plate voltage on the 6L6 was around 362V - This model predicts 364V.

I have no idea what the voltages for the other tubes should be, but based on a schematic for a Supro 1624T I'm thinking around 270V after the last filter cap. The 1624T has a similar pre-amp section with the same tube complement and controls so I'm guessing that is probably a pretty good target. The model predicts 273V.


I could get similar results with a 5V4G rectifier, R1 = 270 ohms.


If my calculations are correct (which I never trust myself)... option 1 drops 52V across a 560R resistor so it would need to be 10W ~= (52*52/560) * 2. Whereas, the second option drops 26V across a 270R resistor so it would need to be 5W ~= (26*26/270) * 2.

Perhaps I didn't do anything wrong before. Maybe I just needed to adjust the values for my transformer. What do you think?

[shooter]

the values sound right, I typically have ~~~470 ohms for my 1st R, which feeds the power tube, 2 more R's in the many k range

[dbishopbliss]

And now, back to our regularly scheduled program… I got the faceplates in.


Years ago I would use the name Slackdog Industries when demoing software. I decided I liked that better than using my last name. Plus, now I get to use my dog’s names for the model.


Time to start soldering.

[sluckey]

Looks top notch!   

[Willabe]

Very nice! 

[dbishopbliss]

I have a wire routing question. I don't think one way versus another would make a difference electronically, but as I found out with my Vox (and reading other's posts) I know that routing wires poorly can make a difference...

The option I see most often is running from the filter cap to the 10K resistor, then from the resistor to screen of the power tube.

For convenience, I would like to route the wire from the cap to pin 6 on V5, then to pin 6 on V4, then to the 10K resistor. There will be fewer wires crossing by doing this and the wires will be shorter. Any reason that I should not hook things up this way?

Note: not all wires are shown for clarity.

[sluckey]

Either way is fine but I prefer the second drawing.

[scstill]

And now, back to our regularly scheduled program… I got the faceplates in.


Years ago I would use the name Slackdog Industries when demoing software. I decided I liked that better than using my last name. Plus, now I get to use my dog’s names for the model.


Time to start soldering.

Where (how) did you do the faceplate? I am building a Supro 1696TN slant cab and am looking to repro the original graphics

[dbishopbliss]

Where (how) did you do the faceplate? I am build a Supro 1696TN slant cab and am looking to repro the original graphics

I designed the faceplate myself using Inkscape. It works pretty well and can read all sorts of file formats including Illustrator, dxf, etc. Mojo shared the dxf template for the JTM 45 chassis with me. I was able to convert that to a layer in the software for the cutouts. I created a new layer for my labels, logos, artwork around that. Then I saved the file as a PDF and sent it to Jeff at JT Productions (phone number is on the website).

The process was very simple and fast. I had faceplates made in the past and there was a lot more back and forth and corrections to the file required (missing fonts, line thickness, etc). I made the process much more difficult because I was worried about that stuff. It seems the new technology is pretty much like using a printer. Just send the PDF, Jeff will send you a version that shows where the cuts will be made and you're done.

If you are fabricating your own chassis without a laser cutter then the challenge will be getting the holes in the chassis to line up exactly with the holes in the faceplate. I ordered faceplates for my AC15 as well (they have to match), but I drilled the chassis by hand so I'm not expecting things to go as smoothly. I used a paper template when drilling the cutouts but it had to be printed out on multiple sheets of paper that were taped together.

If I were to do it again I would order a "pilot hole" template that would only have cutouts around 1/8". Then, mount the template on the chassis with tape. Drill one pilot hole and secure it with a screw. Then drill the rest of the pilot holes. After that, it would be easy to use a step drill bit to drill the rest of the holes to size. That would probably add around $20 to the project but if you are going through the hassle of reproducing the original graphics it will probably be worth the extra $$$.

[dbishopbliss]

It lives! But its only half alive... I say half alive because the voltages seem to be about half of what I think they should be. I was expecting the following:

Point	| Expected	| Measured
PT Secondaries	| 300 VAC	| 316 VAC
Rectifier Pin 4 and 6	| see comment below	| 146 VDC
C1	| 388 VDC	| 197 VDC
C2	| 308 VDC	| 175 VDC
C3	| 278 VDC	| 160 VDC
V4/V5 Plate	| 385 VDC	| 192 VDC

The Plate Voltages were given to me by Leon C so I'm not sure how accurate they will be for my build. The other voltages were based on the PSUD model. Even if they aren't exact, I would expect them to be in the same general ballpark and they seem way off to me.

One thing I did, but I have never done before was install "backup diodes" on the rectifier socket as described by Rob Robinette. They are going from pin 3 to 4 and pin 5 to 6. The PT Secondary wires are connected to pins 3 and 5. The bands of the diodes are on pins 4 and 6. I'm pretty sure that is how things should be wired.


Any thoughts on what would cause the voltages to be so low?

[dbishopbliss]

Here's a picture of the rectifier socket.

[sluckey]

If there are no filter caps connected to the rectifier, the dc voltages will be way low.

[dbishopbliss]

If there are no filter caps connected to the rectifier, the dc voltages will be way low.

Here's another view where you can see the power supply wiring.

There are two red wires connected to Pin 8 of the rectifier; one from the OT and one that goes to the first filter cap.

The 10K voltage dropping resistor is soldered directly to the terminals of the dual 32uF filter cap.

A red wire goes from the second cap to Pin 6 of V5 and V4. I have Screen resistors going from Pin 6 to Pin 4.

A red wire goes from Pin 6 on V5 to the second 10K voltage dropping resistor where you can see the 16uF filter cap is connected (I don't show ground, but its there).

[sluckey]

Remove the power tubes and leave them out until you resolve the B+ issue. Measure resistance from rectifier socket pin 8 to chassis. Should be high resistance.

Are you using a light bulb limiter? If so, remove it. Any better?

Change rectifier tube. Any better?

Use a couple gator clip leads to short the "backup diodes". Any better?

[dbishopbliss]

Remove the power tubes and leave them out until you resolve the B+ issue. Measure resistance from rectifier socket pin 8 to chassis. Should be high resistance.

When I measure Pin 8 of rectifier to ground I get 466K but it continues to climb slowly. Perhaps that's a clue.


I checked the voltages at Pin8/C1, C2 and C3 without the power tubes and measured 430, 420 and 410VDC respectively.

Are you using a light bulb limiter? If so, remove it. Any better?

Change rectifier tube. Any better?

Use a couple gator clip leads to short the "backup diodes". Any better?

No light bulb limiter... I have one but could not find the incandescent bulbs after we moved and my local stores only have LEDs. I only have one 5V4G rectifier... I could swap in a 5AR4, but not sure that is the issue after seeing the higher voltages when the Power Tubes are removed. I didn't try bypassing the diodes for the same reason.

[sluckey]

Power supply seems OK. Put the tubes back in. Measure voltage on pin 5 and 8 of each power tube. What have you?

[dbishopbliss]

Power supply seems OK. Put the tubes back in. Measure voltage on pin 5 and 8 of each power tube. What have you?

I put all the tubes back in, not just the power tubes.

V4, Pin 5 = 0.007 VDC* (Control Grid)
V4, Pin 8 = 13.7 VDC (Cathode)

V5, Pin 5 = 0.008 VDC*
V5, Pin 8 = 13.7 VDC

*Voltages on Pin 5 fluctuated a bit. Measured on my Fluke 77V.

While I was at it I measured additional points:
C3 = 182VDC.
Two 270R resistors come together in a upside down V = 182VDC (it should be because there is an under-board jumper between C3 and those resistors).
V3, Pin 1 = 108VDC (this is connected to the other side of one of the 270R resistors) (plate)
V3, Pin 6 = 107VDC  (this is connected to the other side of one of the 270R resistors) (plate)
After 0.047uF cap = 0.008VDC (connected to V5, pin 5 so makes sense)
After 0.022uF cap = 0.007VDC (connected to V4, pin 5 so makes sense)

I have included a photo of my wiring.

[sluckey]

Put your 5AR4 in and short out the backup diodes.

[dbishopbliss]

Put your 5AR4 in and short out the backup diodes.

I think we are making progress. I swapped in a 5AR4 and measured (with diodes and with diodes bypassed - didn't make a difference).

Point	| Expected	| Measured	| Measured w/5AR4
PT Secondaries	| 300 VAC	| 316 VAC	| 316 VAC
Rectifier Pin 4 and 6	| ?	| 146 VDC	| 311 VDC
C1	| 388 VDC	| 197 VDC	| 391 VDC
C2	| 308 VDC	| 175 VDC	| 337 VDC
C3	| 278 VDC	| 160 VDC	| 309 VDC
V5 Pin3	| 385 VDC	| 192 VDC	| 387 VDC
V5 Pin5	| ?	| .007 VDC	| .016 VDC
V5 Pin8	| ?	| 13.7 VDC	| 25.56 VDC
V4 Pin3	| 385 VDC	| 192 VDC	| 386 VDC
V4 Pin5	| ?	| .008 VDC	| .022 VDC
V4 Pin8	| ?	| 13.7 VDC	| 25.56 VDC
V3 Pin1	| ?	| 108 VDC	| 190.6 VDC
V3 Pin6	| ?	| 107 VDC	| 189.3 VDC

The voltages with the 5AR4 are in the ballpark of what I expected although higher. But, that is to be expected as well since the 5AR4 does not drop as much voltage as the 5V4G

Do you think I have a bad 5V4G tube or could this be caused by using the 32uF filter caps? The datasheet says additional plate impedance may be necessary for 40uF filter caps.

I went ahead and plugged in a speaker/guitar. Seems to sound ok.

[sluckey]

bad tube.

[dbishopbliss]

bad tube.

Ugh! 

[dbishopbliss]

I ordered a 5V4G (two actually) so I will wait until they arrive for an "official" tone reports but I brought the amp up to play with some real speaker cabinets. My bench speaker is a 5-inch 30 watt speaker that is good for proving the amp makes sound, but not much else.


It's pretty quiet. A slight hiss when the volume knobs are turned past noon, but nothing out of the ordinary. No hum or buzz. The channels sound different from each other which is expected; One is cathode biased, the other is grid leak biased. There is not a lot of clean - which seems to be a thing with my amps. :-D I suspect some of that has to do with the higher voltages from using the 5AR4 instead of the 5V4G.


I think I wired the tone knob backwards. Not sure if that is how Supros are ordinarily wired or if I messed up. Regardless, when I turn the knob clockwise it gets darker and loses the highs. I will probably re-wire that so it is more intuitive.


The Tremolo does not work at all. Doesn't matter if I use a foot switch or if I use the push/pull switch on the speed knob so that will need some debugging. I will probably wait until the new rectifier arrives to work on that.


The one thing that is sort of disappointing... it doesn't really sound like the Jimmy Page amp. At least right now. I'm going to wait until I get the proper rectifier to mismatch the load (16 ohm speaker in a 4 ohm output). See how that goes. That said, the amp sounds pretty darn good right now even if I don't sound like Zepplin I.

[shooter]

which seems to be a thing with my amps.

Mine started that way also, the biggest reason in mine, too much gain, slamming the "next" stage
the other was cap values too big, early in the stages

[sluckey]

No need to wait for a new rectifier tube before troubleshooting the tremolo. You did use two insulating bushings on the footswitch jack, right? And you verified that the sleeve lug does not read zero ohms to chassis, right? Disregard my questions about the footswitch jack since you are using a "Cliff" style jack.

I suggest you eliminate the switch and footswitch jack first. This can easily be done by disconnecting the black wire from the FS jack and disconnecting the violet wire from the speed pot switch (referring to your latest layout). Now connect the black and violet wires together and leave like this until the tremolo is working. If the tremolo is already working with the switches bypassed then you only need to examine that wiring. If not, then double check the wiring of the tremolo tube and circuit, paying close attention to the fact that your layout shows using pins 6, 7, 8 for the tube but your schematic shows using pins 1, 2, 3. I've looked closely at your layout in the area that deals with tremolo and see no problems, so if your amp is really wired IAW your layout, the tremolo should work. Post the voltage readings for pins 6, 7, and 8.

[dbishopbliss]

Not sure what I was thinking but I totally botched the foot switch. I think it’s because I made the faceplate match the original control layout without double checking my component layout. So I had to swap the pots and then I rotated the jack but when I wired I didn’t look at what the lugs do, just followed the wires.

[sluckey]

So, is the tremolo working now?

[dbishopbliss]

I just fixed the wiring and no luck.


I’m using switchcraft jacks now because I wanted them all to match and could not did a cliff equivalent of the 13A.


To answer your question, sleeve to ground is 67K with the pot switch in, 98K with it out.


With the pot pushed in
V2, 6 = 229
V2, 7 = 0
V2, 8 = 2.7



With the pot pulled out
V2, 6 = 72
V2, 7 = -0.484 (negative???)
V2, 8 = 0.1


I will have to try connecting black and violet wires later on. I have other commitments now. Do I need to disconnect both wires or could I only disconnect the black wire and then use clips to connect it to the violet wire in place? The way I installed the ground bus makes it a little tight.

[sluckey]

I’m using switchcraft jacks now because I wanted them all to match and could not did a cliff equivalent of the 13A.

So, did you use two isolating bushings on the F/S jack?

To answer your question, sleeve to ground is 67K with the pot switch in, 98K with it out.

Sleeve to ground should be 3.9K and should not change when you mess with the pot switch. This 3.9K resistance is provided by the 3.9K resistor on V1 pin 8.

I will have to try connecting black and violet wires later on. I have other commitments now. Do I need to disconnect both wires or could I only disconnect the black wire and then use clips to connect it to the violet wire in place? The way I installed the ground bus makes it a little tight.

I was very clear about those wires.

[dbishopbliss]

So, did you use two isolating bushings on the F/S jack?

Yes. When I use the continuity checker on my meter I get no sound.

Sleeve to ground should be 3.9K and should not change when you mess with the pot switch. This 3.9K resistance is provided by the 3.9K resistor on V1 pin 8.

I disconnected the black wire and it measures 3.9K as it should. Maybe that's a a clue.

I connected the black and violet wires (its actually yellow in real life, but we can keep calling it violet for clarity)... still no tremolo. You only asked for the voltages on V2, Pins 6, 7 and 8 but I have included all the voltages for V1 and V2:

V1, 1 = 92V
V1, 2 = -0.833V
V1, 3 = 0V
V1, 6 = 193V
V1, 7 = 0V
V1, 8 = 2.27V

V2, 1 = 48V
V2, 2 = -0.15V
V2, 3 = 0.513V
V2, 6 = 201V
V2, 7 = 0V
V2, 8 = 2.27V

[sluckey]

Sleeve to ground should be 3.9K and should not change when you mess with the pot switch. This 3.9K resistance is provided by the 3.9K resistor on V1 pin 8.

I disconnected the black wire and it measures 3.9K as it should. Maybe that's a a clue.

The black wire should read 3.9K because the other end is connected to a 3.9K cathode resistor. So, it's incredible to have "sleeve to ground is 67K with the pot switch in, 98K with it out". What resistance do you have between the sleeve and chassis while the black and violet wires are conned together?

V2, 6 = 201V

Is this voltage steady or is it bouncing around (would look erratic on a DMM)? And what is the voltage at the Node C filter cap?

[dbishopbliss]

With the black and violet wires connected I did not get any reading between sleeve and chassis. Just to be sure I removed the jack from the chassis so it is floating in air. There is no way it can be making a ground connection now. Still no Tremolo.


The voltage on V2,6 is steady although 200V this time. I took voltages at various points throughout the circuit and annotated the layout in the attachment.

[sluckey]

With the black and violet wires connected I did not get any reading between sleeve and chassis.

And that's what it should be. So, your statement "sleeve to ground is 67K with the pot switch in, 98K with it out" is still a mystery.  That's OK. Reinstall the jack and connect the black and violet wires back as they should be and lets start over.

The steady plate voltage means the trem oscillator is not oscillating. Set the INT pot to mid rotation and be sure nothing is plugged into the F/S jack. With amp off, measure resistance from pin 8 to ground. It should be 18.9K with the pot switch off and 3.9K with the pot switch on.

Turn the amp on and measure the voltage on V2 pins 6, 7, and 8 with the pot switch off. Now turn the pot switch on and repeat voltage measurements for V2 pins 6, 7, and 8. Also verify that V2 pin 6 voltages changes as you turn the INT pot.

Also post some hi-rez pics of the guts. I'm mostly interested in seeing V2 and the portion of the board with the tremolo circuit and connecting wires. It'll be good if I can see board jumpers too.

[dbishopbliss]

That's OK. Reinstall the jack and connect the black and violet wires back as they should be and lets start over.

I added extra insulation (clear enamel) to the jack and the chassis just in case. Now when I reinstall the jack the there is a 3.9K no matter how things are switched. That seems to be sorted out, but still no Tremolo.

The steady plate voltage means the trem oscillator is not oscillating. Set the INT pot to mid rotation and be sure nothing is plugged into the F/S jack. With amp off, measure resistance from pin 8 to ground. It should be 18.9K with the pot switch off and 3.9K with the pot switch on.

18.86K and 3.9K - close enough.

Turn the amp on and measure the voltage on V2 pins 6, 7, and 8 with the pot switch off. Now turn the pot switch on and repeat voltage measurements for V2 pins 6, 7, and 8. Also verify that V2 pin 6 voltages changes as you turn the INT pot.

V2,6 Off = 265
V2,7 Off = 0
V2,8 Off = 3.39

V2,6 On = 204
V2,7 On = 0
V2,8 On = 2.28

Also post some hi-rez pics of the guts. I'm mostly interested in seeing V2 and the portion of the board with the tremolo circuit and connecting wires. It'll be good if I can see board jumpers too.

Oops... wrong images. hard to tell from the thumbnails on my computer.

[dbishopbliss]

Here is the pots and jack - yeah should have routed that black wire better. :-D

[dbishopbliss]

V2 Socket... nothing to see here.

[dbishopbliss]

turret board components.


The red jumpers are on top because I accidentally soldered the wire in the wrong turret when I was first running the jumpers. Because the wire was too short, I had to cut a new wire but then the solder that was left in the turret was keeping me from being able to install the jumper under the board again so I just ran them on top.


Red wire from the intensity pot is connected under the board to the turret where the other red wires are connected.


Blue wire from intensity pot is connected under the board where the 270K and 100K resistors come together.


All other jumpers are on top of the board.

[sluckey]

I really want to see one pic that shows the board and V2 socket together.

[dbishopbliss]

Here you go

[dbishopbliss]

I was just reviewing Valve Wizard's write-up of The Phase-Shift Oscillator For Tremolo.


It seems like all the parts are there but there are a few differences not described in his page:


1. The intensity pot... I'm not quite sure how to calculate Ra. Would it be 270K or would it be 270K + (500K || 150K).
2. He uses the same values for all the caps while the 1690T has different caps. I can't imagine that is the issue.
3. The resistors for R1, R2, R3 are all quite different that what he suggests.
4. This is the biggest difference... he shows the foot switch being between C2 and C3 while the 1690T has it off the cathode. I haven't compared to other amp schematics yet. Could the schematic be wrong?


I keep re-tracing the circuit and I'm not seeing any other issues. By the way, I swapped V2 for a different tube. Just in case that was the issue. Nope... same behavior. No Tremolo.

I also had the sleeve and shunt connected which the schematic doesn’t so I removed that. Everything is connected to the sleeve only now. Still no tremolo.

[sluckey]

I also had the sleeve and shunt connected which the schematic doesn’t so I removed that. Everything is connected to the sleeve only now. Still no tremolo.

The schematic clearly shows the sleeve and shunt connected. The circuit cannot possibly work with that connection removed.

[dbishopbliss]

The schematic clearly shows the sleeve and shunt connected. The circuit cannot possibly work with that connection removed.

You are right (of course). I got myself confused trying to find something I missed. I put it back. No difference.