GROUP BUY: Universal Audio 176 - a 100% faithful recreation of a legend! CLOSED!

[Potato Cakes]

I pulled it apart and added the extra ground wires to both PCBs. I found one spot on the bottom board that seemed to have a standoff shorting to an adjacent trace. I also found that the outer rim of the PCB has a copper trace that I believe was left over from the etching process which was shoring to an actual trace which was very close by. I cut between them and verified there was no longer a short. I checked to make sure the connections were being made from each solder point the reassembled it. I still have the the stupid buzz and 2.3VAC on power up. With the unit powered off but still plugged in I am getting 0.43VAC between pin 3 and ground. Also, when listening to the unit, if I flex the bottom PCB enough the buzz lessens considerably but never goes away. I also managed to lay the soldering iron on my middle knuckle and experience my first tube voltage shock as my frustration and anger became more visible. I'm quite bummed out at the moment as I am having hard time understanding where this voltage to ground it coming from and the fact that unless I spot the issue or get a better idea of what could be happening I will then have to start pulling it apart again. The only thing that comes to mind is a bad power transformer or one of the can caps has an issue. Or the PCB in question has more problems with it.

In the interim, I'm going to redo the grounding on the one that is much quieter and see if I can get rid the buzz all together.

Thanks!

Paul

 

[Potato Cakes]

All righty, one is finally done and in the rack. I took the less noisy of the two and rewired the grounding scheme. I ran individual wires to both of the can caps mounted on the chassis. the meter -, ratio resistor ground, and attack switch ground were all tied together and ran to the star ground as one unit. On the bottom PCB, the bottom ground connection was already connecting to the chassis somehow, and I had no desire to rip apart this one that was already working well, I just left it and connected the top ground connection. I think it made a little bit of a difference, but I didn't take a before measurement as I was so focused on getting at least one and I was pretty angry at that point about some additional setbacks I had in ground rewiring process. But I did take some afterwards and I am measuring about a 95dB S/N ratio at 1kHz with the following settings:

Input: 20
Output: 18
Compressor: off
Both Venier controls all the way down
1kHz tone at -20dBFS (this measures 0.775VAC on the output of my converter)

The buzz is still there if I really crank on the output control, but it's in a region that would be unusable for me regarding gain structure. In initial listening tests with the controls set to how I would use this guy for various artistic decisions, I cannot hear any noise. Even when stopping the playback, I have to strain to hear what I think might be the buzzing sound I was describing in the previous post. I'd be curious to know what adding a 6.3VDC PSU to use for tube heaters would do in terms of noise. At any rate, what I am getting is well within the acceptable range of modern tube gear expectations.

Here is a list of modern substitutes that I used in place of parts that may be very hard to near impossible to find and within a reasonable price:

- Input/Output attenuators - Bourns T-pads from CAPI
- C6/C7 - Axial Film 0.47uF 630V - I can't remember the brand. They are yellow and possibly Vishay. If you really had your heart set on some oil can caps but couldn't find the right size, you can always spend the extra money for the high end paper oil ones and still feel fancy
- PRP resistors through out except for the 1M2 resistors. Those are carbon comp. Don't ask. That's just what happened.
- Output transformer - Sowter 1290e. I think this is currently the only readily available option. Don Audio stocks them (or did) so you don't have wait for the 3 week lead time directly from Sowter
- Test switch - Lorlin 2x6 - The layout has a 3x4 switch but the third pole isn't used, so I went with a Lorlin 2x6 that I had.
- Venier controls - 500R pot with 750R parallel. This seems to work just fine.
- Power transformer - In the US, a company called Musical Power Supplies makes a power transformer to the UA176 specs as jslstudio pointed out. These work great and are less cost prohibitive than the Hammond option.

I also left out the internal meter switching as I won't have any need or desire to open the unit up just so I can look at the output level. To do this, run a wire from the GR Zero wiper pot on the top PCB to the Meter + connection. Meter - goes to ground.

For calibration, I had to make something up based on some poorly educated guessing, but the results seems to work well. I set the compression to 12:1, injected a 1kHz tone and adjusted the input till I got about 10dB of reduction on the meter. I then measured VDC on CW and CCW for both plate and cathode balance referenced to ground and made adjustment so that both sides of each pot had the same voltage. I'm not sure if I would have achieved the same results with no signal present but I didn't bother to check. If there is a better way do this with just the GR function on the meter then I'm willing to try it. Currently I'm very pleased with how this build performs with my faux calibration.

I would have liked to have built one with all original parts, but the end result still sounds really great and I would be curious to do a comparison with a "real" one if I could find one in Nashville to rent or have someone bring by. The parts used are per schematic values and have much tighter component tolerances than when these were made 50 years ago, so I have no doubt this will perform as well as a good working sample of originals. If you do lose patience with trying to make these 100% historic, you should know that using modern alternatives will still give you the sonic and compression characteristics that made you want to build one of these in the first place.

I've attached a pic of the measurement taken with 1kHz tone.

 

[Potato Cakes]

And here is the baseline with no signal present and the same control settings. Converter used was a Lynx Aurora set to 24/192.

For whatever reason I get a weird reading in the low end with no signal present using this plug in. I definitely don't hear that low end bump in noise.

Now to figure out what is going on with the other one...


Thanks!

Paul

 

[Phrazemaster]

Manual for 176

https://www.sowter.co.uk/schematics/UA17567Litcomp.pdf

 

[Potato Cakes]

I believe I have a copy of this downloaded. The procedure described therein requires the metering selection to monitor output. My takeaway from reading it is balancing the plate and cathode is balanced voltages, which is what I am doing (I think) while the components are under signal load and compressing. I think I'm just going to go with this for now as I said it is performing like one would expect and I don't hear any undue distortion or frequency loss. I'll do further measurements later to confirm this.

Thanks!

Paul

 

[Phrazemaster]

My unit has been working perfectly ever since - even though I used 500ohm pots with a resistor in parallel for the verniers, so your choice of pots there is not as crucial as long as you use good quality pots.
I eventually replaced my "regular" vernier pots with switchable ones as I modded my unit along the way and it still works great.
I used push-pull switched pots for both verniers - the input one switches the unit to hi-gain when pulled (optional as per the original pcb) and the output vernier switches the meter from GR to OUTPUT when pulled...

Hey Martin, just wondering which switched pots you used? Those mods sound ace!

Thx,

Mike

 

[Potato Cakes]

I pulled the second one apart and copied the grounding as before. I scoured both main PCBs and checked again for shorts and that connecting traces in fact had continuity. I found that I was missing a connection between C10 and R38 on the PSU and that a lead was causing an unintended connection on one of the front jacks, but neither of those proved to be the culprits. I have been working on this nonstop for 9 hours today and what is causing the buzz is escapes me still. It sounds like a grounding problem, which at this point is almost impossible unless I have a bad component.

I did go through tubes. With only the 12BH7 pulled the noise goes away. I swapped this and both power tubes with the other unit with but it did not solve the issue. With all the other audio tubes pulled and just the 12BH7 installed the noise is still present. Hopefully I'm on to something and not chasing the wrong thing because I misinterpreted the schematic.

Thanks!

Paul

 

[Potato Cakes]

Alrighty, some helpful clues have emerged.

Without the 12BH7  it is not completely quiet. It is however this way in the working unit. In the noisy unit, when the 12BH7's are pulled, I get about the same noise level that I should be getting to match the other 176. I decided to listen to it again to verify they at least sounded the same with equivalent settings and there was a drastic difference in the compression characteristics. At 12:1, the working 176 still sounds smooth even at high attack settings, while the noisy one has significantly more gain reduction and the attack, even at a slower setting, is much more aggressive. Also, the output is not only less on the noisy unit but seems slightly band passed as the working one has better high and low frequency response. This is possibly an artifact of the more exaggerated compression characteristics. I swapped the 6AL5 tubes but fix the problem. I am getting 38VDC for pins 1 and 5, which are 7V less than what is listed on the schematic, but this is the same value (I think) that I got on the other unit. I'll have to double check that. The values and connections that I have on this build are lining up with the schematic when doing continuity tests, so even though I think I've narrowed it down I still can't seem to find what ever is ailing this guy. I did the same plate and cathode balance procedure as on the working unit, which were initially off and had even more aggressive compression characteristics. The heater voltage is 6.6VAC at the tube.

Any thoughts? I know it's something stupid as it usually is, but I think my problem is that I've stared at the same schematic and layout for so long and rewired this thing so many times my eyes are starting to glaze over and it's getting hard to come up with fresh new approaches.

Thanks!

Paul

 

[Potato Cakes]

At it again tonight. I pulled it apart again. I found that where the R23 and R24 connect their solder legs poked through and seemingly melted through the covering of C1 on the other side. This would seem to make sense about I was experiencing. It did not seem to go deep enough to damage it, so I put some electrical tape on it temporarily and swapped position with C2, but this turned out to be a bust as well. I angrily started probing around to see if could make the noise go away and still pass audio and managed to get shocked a second time trying to fix this stupid noise. It took every bit of self control not to get my 4lb hammer and beat this into a folded piece of scrap metal. The problem is my OCD won't let me walk away from this. It's a buzz and the compression is much more aggressive and erratic than the working unit, and as much pulling apart I've done and redoing everything I should have found the problem.

I've attached a pic of the noise I am experiencing with approximately the same level settings as the working unit in case someone may have seen this before during their troubleshooting process.

Thanks!

Paul

 

[core13]

what are you using for the 0,047 C3 & C4 radial caps?

it's hard to find radial polypro caps exept orange drop

 

[Potato Cakes]

I'm actually using axial caps installed like resistors when building a 2520 style op amp. The lead that is exposed I cover with shrink tubing for safety.

Thanks!

Paul

 

[core13]

thanks

it's I wanted to do, but curious if someone know the original model of C3 C4 caps?

 

[Potato Cakes]

The second one is now done!

After rebuilding it for the 5th time and still getting the same noise, I installed a DC tube heater that I built from one of the boards currently being sold on eBay which are really well designed. I installed it for the input, output and compression tubes. I left the GZ34 and OB2 connected to other 6.3VAC winding. I fired it up and it is even quieter than the other unit. There is no hum or buzz, even when I crank the output of the 176 all the way up and then turn up the monitoring section. With every thing wide open and I hear some faint fizzling/crackling which I can see on the same analyzer plugin as before, but in normal operation it is dead quiet. I believe this from one of the tubes.

I also discovered that the super aggressive compression was from the 6BC8. Before I installed the heater, I was still trying to figure out where the noise was coming from as, like I mentioned the before, the other unit was built with AC heaters per the layout/schematic and is quiet. I swapped all of the tubes from one unit to the other and compression characteristics moved with it, but not the noise, making me realize that one had nothing to do with the other and that I had been using this line of thinking for troubleshooting in error. Started swapping them back and when I did the 6BC8 the compression characteristics swapped back over. The spare one that I ordered turned out to be dead, and I also managed to break a 6AL5 when swapping everything back and forth several times.

What are some recommendations for either of these tubes that function well and are fairly consistent between batches?

Thanks!

Paul

 

[Potato Cakes]

For ensuring rivets are indeed making contact with the traces, I found a good thing to do before you install components is to apply some solder on the trace connecting to the rivet at a very short distance away and work to the rivet till you see solder flow around it but not in the hole. Then when you solder in components they are making direct connections to the traces and thus the other parts which they need to make an electrical connection. It seems like a bit of a pain, but it will save you time metering between all the solder points trying to find a missing connection, which may not be visually obvious.

Thanks!

Paul

 

[rainton]

The second one is now done!

After rebuilding it for the 5th time and still getting the same noise, I installed a DC tube heater that I built from one of the boards currently being sold on eBay which are really well designed. I installed it for the input, output and compression tubes. I left the GZ34 and OB2 connected to other 6.3VAC winding. I fired it up and it is even quieter than the other unit. There is no hum or buzz, even when I crank the output of the 176 all the way up and then turn up the monitoring section. With every thing wide open and I hear some faint fizzling/crackling which I can see on the same analyzer plugin as before, but in normal operation it is dead quiet. I believe this from one of the tubes.

I also discovered that the super aggressive compression was from the 6BC8. Before I installed the heater, I was still trying to figure out where the noise was coming from as, like I mentioned the before, the other unit was built with AC heaters per the layout/schematic and is quiet. I swapped all of the tubes from one unit to the other and compression characteristics moved with it, but not the noise, making me realize that one had nothing to do with the other and that I had been using this line of thinking for troubleshooting in error. Started swapping them back and when I did the 6BC8 the compression characteristics swapped back over. The spare one that I ordered turned out to be dead, and I also managed to break a 6AL5 when swapping everything back and forth several times.

What are some recommendations for either of these tubes that function well and are fairly consistent between batches?

Thanks!

Paul

glad you made it Paul!

I was on the road the entire week, but when I read about the weird compression behavior I was about to say that it's the 6BC8 since that tube is responsible for compression in the 176.
That's the "variety-mu" tube. I had a bad one too when I first fired up my prototype. The good thing is that they're not expensive, so I just bought a bunch of them - mostly RCA.

In terms of noise coming from the AC heaters routing and twisting the wires tightly is crucial. But I'm sure you know all of that.
I just remember when restoring an old Ampeg B15 I also had a buzz going on and in the end it were several minor things I changed in terms of routing  that solved the problem.
You did triple check all ground wires do have a proper connection I assume? Not only that you multimeter shows connection but also the resistance it's measuring when checking for continuity...e.g. in one build I removed the star ground screw from the chassis and removed a little bit more of the chassis' powder coating around that spot and it helped tremendously.

Anyway - glad you made it!!

 

[Potato Cakes]

I believe I had the heater wires from the transformer twisted well enough. Again, after the first build was successful, I copied that one while trying to make the wiring neater, so when it didn't have the same low noise as before the madness ensued. I don't know how many more times I could have checked ground connections. And I did take apart the star ground connection and remove more paint from the area and tighten it back up to the extreme. I do know what you are talking about with testing continuity and getting 1-5 ohms of resistance as I had that happen when checking between the points on the board. I have a Fluke 112 that I've been using for years, and the only time I've had issues is when the test leads start to go bad. So the only thing I didn't redo was the twisting of the heater winding from the transformer, and that very well may have been it. But I'm not going to go back and verify as I have to move on to other projects. It may not be completely historic, but it is not uncommon for studios to have modified their old tube gear to deal with noise issues, so maybe it could be considered historic in that sense.

Thanks again (again) for making this project available. Much was learned about this circuit and I now have two amazing sounding compressors to use on an upcoming project.

Thanks!

Paul

 

[Potato Cakes]

One more follow up. I installed a DC heater circuit in the unit that was already quiet. I did notice that I had a tighter twist on those heater lines than I did on the noisy unit, which may have been the culprit all along. With the heaters running on DC, all remnants of buzz and hum are completely removed. With the monitoring and output of the 176 cranked all the way up, I only hear a slight amount of hiss, which is what I would expect at those extremes. It's about the same as a good solid state circuit. In normal use, I don't hear anything noise wise. It's almost as if the monitoring is muted. I highly recommend adding this to one's 176 build, especially if they're in an area where the local power has a tendency to cause noise.

Thanks!

Paul

 

[BramK]

I've enjoyed following the builds here. Glad that things are finally coming together Potato Cakes.

I'm getting ready to start my build. I ordered most of the components way back when. I quickly perused my stock this evening and had a few questions come up.

1. 372X mounting holes don't fit the chassis cutout vertically. Two holes will line up, but not all 4. Are folks just mounting the top two screws?

2. C6 & C7 - Looks like I ordered some old PIO Gudeman caps, they're axial ones, not the square cans. Mine are rated 400V - will that be sufficient or is 600V needed here?

3. Resistors - I'm not concerned with visual originality here, but I want to do good on the character sound. I went almost entirely metal film with my resistor purchases last year. Which spots in this design would benefit sonically from the Carbon Comp distortion?


4. What is a recommended order to approach this build? Mount chassis back panel transformers/sockets/hardware first? Populate PCBs? Wiring?

Thanks!

 

[moltenwok]

Hi team, building up all 4 at once here, with the 372x mounting I just milled out the mounting holes in the transformer feet and put big pan head screws in. That way didn't have to drill chassis or change pcb mounts. Cheers

 

[Potato Cakes]

I've enjoyed following the builds here. Glad that things are finally coming together Potato Cakes.

I'm getting ready to start my build. I ordered most of the components way back when. I quickly perused my stock this evening and had a few questions come up.

1. 372X mounting holes don't fit the chassis cutout vertically. Two holes will line up, but not all 4. Are folks just mounting the top two screws?

2. C6 & C7 - Looks like I ordered some old PIO Gudeman caps, they're axial ones, not the square cans. Mine are rated 400V - will that be sufficient or is 600V needed here?

3. Resistors - I'm not concerned with visual originality here, but I want to do good on the character sound. I went almost entirely metal film with my resistor purchases last year. Which spots in this design would benefit sonically from the Carbon Comp distortion?


4. What is a recommended order to approach this build? Mount chassis back panel transformers/sockets/hardware first? Populate PCBs? Wiring?

Thanks!

Some suggested answers to your inquiries:

1 - see moltenwork's answer

2 - On the schematic, it shows a test voltage of 310V that goes from R44 on the PSU through both primary windings and connects to C6 and 7. Using modern components this might not be an issue, but whether you are using NOS or new parts it would be safe to use 630V rated ones.

3 - If you read page one of the 175b manual posted by Martin from Sowter's website, it says in the first paragraph that the unit was designed for low distortion to be used in high quality audio systems and in the second paragraph that it was designed with overrated components and precision 1% resistors. It seems that the whole series of these limiters (175x, 176) were designed for fidelity and not tone (distortion). I wouldn't get too hung up on where to install carbon comp for character. When you use this unit with fast attack and release settings you naturally start to get distortion/artifacts/character/tone or similar quality. Also, can you even buy 1% carbon comp resistors? I've never seen anything below 5%.

4 - As a general rule, I would start on the PSU side and work the left as you look at the chassis from the front. Here's my recommended order of operations:

1. Build PCBs (this can be done at any point up to actually installing them). I would strongly recommend before you install any components that you solder any and all rivets where they connect to their respective traces. Apply solder directly to the trace and then work it to the rivet. With your iron touching the side of the rivet, leave it and work your solder on the side of the rivet till you see it flow around it and adhere to it's trace. Obviously you don't want it to go inside of the rivet. Do this for every one, even if there will be no component placed there as some of the rivets are vias to make connections on the other side of the board.

2. Install Power transformer. Get a hand drill to do the twisted pairs per the build guide. Grip the leads being twisted at the transformer so you don't rip out the windings. Use the slowest setting, naturally.

3. Install the inlet AC cable, GZ34 socket, OB2 socket. Wire the 325V leads to the GZ34 socket. Any connections on the OB2 and GZ34 should be made with appropriate wire lengths to their destination per the wiring guide. You want this whole section prewired as you won't be able to get to them when the choke and PSU board is installed.

4. PSU board - I would recommend making your connections between components on the PSU board on the back using the appropriate turrets.

5. Install 6AL5 socket and C6,7 if you are using the designated mounting holes.

6. Install Input, Interstage, and output transformers. If any of these are original you need to prewire these as well.

7. Barrier strip. I put shrink tubing over each terminal where it went through to chassis to ensure no contact was being made there.

8. Input/Output jack assembly. Prewire this as well and double check to make sure the terminals for passing signal do indeed have continuity. Same for the ground

9. Install front panel. You could do this before the jacks if you wanted. One of the nuts is under the jack assembly which you may find difficult to tighten with the jacks in place. For me it didn't matter, so 8 and 9 are which ever you prefer.

10. Install PCBs. I would recommend using standoffs that are slightly varying in length to allow the appropriate leads from the output transformer to make connections to the ratio switch buy passing between the top and bottom PCBs. Otherwise you will have to extend them to reach with the front panel fully open.

11. Install front panel components.

12. Install tubes.

13. Calibrate/Test. Be sure to have several 6BC8 tubes as I've had 2 out of 7 test/perform poorly. A good gauge is to set the ratio to 12:1 and attack/release to 12 o'clock and run some known program material that hasn't been smashed to pieces to be earbud optimized. Bad tubes will compress (if they even pass sound) in a dramatic fashion on the transients with extreme swings in gain reduction, almost as if there is an intermittent audio connection. I'm sure someone here with a tube tester can give you objective measurements for what range to expect in a quality functioning 6C8B tube.

These are my recommendations and in no way any sort of official guide. Use at your discretion/risk. Let us know how you get on.

Thanks!

Paul