C11 Preamp Restoration - Advice requested on tweaking voltages

[sabrefencer]

I recently acquired what appears to be an original C11. It has the silicon rectifier bridge (either the mod per service bulletin 105 or factory) but retained the original selenium rectifiers for the B+. Unit appears to have been produced in second half of 1962 (based on various component date codes). The unit was in working condition when received, although one channel was intermittent and had lower output. Unit has 6 original Telefunken smooth plate 12AX7, the 12 triode sections test between 66% and 90% of the bogey on my Hickok 800A, no shorts, no gas.

Before doing any work, I took voltage readings per the factory service manual - more on this below. I also checked for DC at the outputs, there was none, and scoped the outputs. Using the side output levelers, I was able to balance the output voltage, but one channel looked terrible with a square wave. I then cleaned all the pots, switches, and tube sockets with a sequence that included CRC QD Contact Cleaner, Deoxit5, and F5 fader lube for pots. Cleaning was done with faceplate and glass removed to ensure these pretties did not get damaged from spray. The cleaning worked wonders - see before/after scope photo.

Now I was ready to start the recap - focusing on the power supply, I installed three new can caps (CE 50/50/50/50 @350v, CE 40/20/20/20 525v, and Authenticap 2000/2000 @35v) + discrete caps for C104 (47 @350v) and C102A (1000 @40v). The "kit" I used also included a 20v Zener, and 1N4007s to replace the selenium rectifiers, and two 1N5408 to adjust heater voltage. I did not initially install the 1N5408s. I also replaced power supply resistors R101 through R106 with new 1/2W metal film resistors, since I was in there.

After the cap install, I checked voltages at 117 and 122 vAC. See voltage chart.

With the Zener installed between C104 and R102, the voltages were between 8% and 11% below spec at 117v (very close to the pre-recap values) except for the heater voltage, which was a about 7% high (higher than before the recap). I then removed the Zener from the circuit (returning to stock except for 1N4007s), this brought the B+ voltages up to within 3%- 6% below factory at 117v, not bad, but the heaters were still above 20v, vs. spec of 18.9. To see if I could get the heater voltage down I put the two 1N5408s in series between C101A (1000 uF) and the heaters (connection to the string is at V6). This brought the heater voltage down about 1v to 19.06, which is less than 1% above spec. I then increased the line voltage to 122.1 to see what voltages look like with my normal line voltage, this pushed the B+ voltages to within +/- 2.5% of spec, and the heater voltage went up a full volt to 20.15v.

The voltage directly out of the bridge is also higher than spec, it is running about 30v vs. 27.6, could I bring this down with a resistor? I can run this through a decade resistor box to get the right resistance rating, would a 1W do, or should I go with a higher wattage. Bringing this voltage down may give me the heater voltages I am looking for without the need for the 1N5048s.

With all these permutations, I am trying to figure out the "best" compromise, and also, why with new caps and resistors, can't I "match" the original voltage specs?

If a resistor after the bridge does not bring the heater voltage down, I will likely go with the 1N5408s in series, but I am waffling between lower B+ voltages with the Zener in place, and closer to spec, but still slightly low @117, voltages without the Zener.

So the question is would you keep the Zener? What is the downside to running low vs. slightly high B+ voltages?

Thanks for your input - Gary

 

[sabrefencer]

Before getting started I did my research and found three notable threads, two AK threads, one from member DC in 2014 who did a very nice job on a C11, and one from member Analog Addict who did a nice job on a C22 (nearly identical circuit), I also looked at a thread from NOS valves. The work in these threads also included replacement of signal caps, which I plan on doing, but I want to get the power supply stable before I attack the signal path. Here are the links to to the threads:
DC: https://www.audiokarma.org/forums/i...y-bucket-list-mcintosh-preamp-c11-c22.583360/
Analog Addict: https://www.audiokarma.org/forums/i...a-c22-restoration.932311/page-2#post-14704914
NOS Valves: http://www.nosvalves.com/C22.htm

 

[sabrefencer]

After discussing with Carmine at ctech, I decided the voltages with the zener were within tolerance, although just slightly low. Two 1N5408 installed before heaters. Attached are the voltages I ended up with. Will now do some more testing, listening, and then re-cap the signal path. Will post more photos, when I get to that, which will likely be next month.

 

[dcgillespie]

Gary -- I would not get too hung up on achieving exact schematic voltages, as there are too many detail variables involved to achieve that at every node, and also, vacuum tubes operate quite well over a wide range of operating voltages. Still, working to achieve good agreement with the schematic is a good indicator of proper operation.

When replacing selenium rectifiers with silicon, the (imo) poor guidance out there regarding this topic is seemingly limitless, and can result in various degrees of success in mimicking the performance of the original selenium (or not). More specifically then:

1. DC Heater Supplies: In converting to silicon, even some manufacturers did a hasty job of conversion, creating issues they either missed, or chose to ignore. Within vacuum tube audio equipment, these are low voltage high current supplies. The high current causes a voltage drop across the selenium device, such that when silicon is installed, their lower voltage drop causes voltages rise, with the added diodes offered then installed to bring operating voltages back to normal.

What this approach fails to address is that with the installation of silicon rectifiers, their very low conduction resistance causes the peak current draw from the power transformer heater winding to be greatly elevated due to the increased charging current now flowing through the first filter cap. Output voltage to the load rises, which the additional series diodes address, but what's left unaddressed is the increased transformer heating from the increased peak current draw such a conversion causes. The ripple current the first filter cap must handle is also increased, impacting its life as well.

To address these concerns, the approach I use is to not use the additional series diodes but rather, install (typically) a 2Ω 5W resistor between the output of the new silicon bridge, and the first filter cap. This resolve then really does address all the issues of converting selenium to silicon in high current power supplies, by augmenting the low internal resistance of the new silicon devices with some additional external resistance. This then allow the new silicon devices to closely mimic the performance of the original selenium device when new. As a result, transformer heating and ripple current remain the same as in the original design, as does output voltage as well. Other than the heat given off by the resistor (replacing the heat produced by the original selenium device), this is usually a much better approach to use since it addresses all the issues involved. The resistor may need some slight adjustment in various applications, but the value given is typically pretty well right in the ball park in most cases, and what I used when I restored my own C11s.

2. B+ Power Supplies: Here -- being a preamp design -- the peak and load current flow through the rectifiers is so much less in comparison to the previous scenario, that the voltage drop produced by good selenium devices versus that of silicon is not that much greater than you might think. As a result, using a voltage adjustment scheme employing a Zener diode will often produce a lower operating voltage than specified for the load -- just as you experienced.

The best resolve then is exactly same: Replace the selenium with silicon, and then add resistance as appropriate in series with (in the case of the C11) the power transformer HV winding lead that connects to the voltage doubler rectifiers. Again, this addresses any over voltage concerns, as well as the increased peak current draw that would otherwise be drawn by both doubler caps as well. In my C11s, I used a 47Ω 0.5 watt resistor to compensate for the installation of silicon devices.

Remember too that in the C11, the B+ supply operates in series with the DC Heater supply. If this relationship is not maintained, then the B+ will start out lower by about 28 vdc to begin with.

I hope this helps!

Dave

 

[sabrefencer]

Dave,
This helps immensely! I was not comfortable with the way it was. I have translated your description of the heater and B+ circuit corrections into the attached annotated photo and schematic. Unless I made an error, this is how I will implement your recommendations.

It looks like the bridge rectifier for the heaters was added after the unit left the factory, the diodes test OK, but it is kind of ugly. Would you leave it, or replace it with a modern bridge? I have a 3A and an 8A bridge on hand, would the 3A be enough, or would you use the 8A?

Diagram has been updated, per input from Dave G.

Gary

 

[dcgillespie]

Note too that in your annotation, you also need to note moving R107 to connect after the new 2Ω resistor as well. Otherwise, it looks good.

The load draws 0.6 A, and when you include ripple current, peak current draw through the bridge won't be anything significantly greater than 1A, so a 3A device is plenty, heat sunk to the chassis if possible. While a 100 volt device would work, using at least a 200 volt device will ensure that no everyday line spikes will cause any damage to it.

Dave

 

[sabrefencer]

Dave,

Good catch on the EXACT location in the unit of the new 2Ω resistor. I have revised the notation on the diagram, and added ID for R107. It is hard to see the lead to R107 in the photo, but it attaches to the same lug on the rectifier as the orange wire. I will likely create a new tie point for the orange wire and the lead to R107, and then put the 2Ω resistor between the bridge and the new tie point.

I will be changing out the bridge with a 3A 400PRV 4-square unit screwed to the chassis using the screw for existing the bridge lug strips so I don't have to drill any new holes. Is heat-sink compound between the bridge and the chassis a good idea?

Gary

 

[dcgillespie]

Always.

 

[sabrefencer]

Looking to order the 2ohm 5W resistor, most in that value are wirewound, only found one that was not. Is wirewound OK?

 

[dcgillespie]

Absolutely.

 

[sabrefencer]

Back from vacation and revised the power supply as described above. I replaced the 4 lug strip with a 5 lug strip (4+ground) to create an anchor/tie point for the 2Ω 5w resistor, R107 and the orange wire feeding the C103A. Lug strip uses original screw from one of the selenium diodes, so no holes needed to be drilled - and no animals were harmed. 3A 400PIV bridge rectifier is mounted using existing hole for original discrete diode bridge. New bridge is seated with Dow 340 heat sink compound.
B+ voltages came up slightly and are now within 2% - 6% of spec (on the low side) with line voltage at 122v. Heaters are nearly spot on at 122v.

Dave, a huge THANK YOU for your help.

New voltages below, and photos attached.

With the power supply where it should be, the next step is to re-cap the signal path.

 

[dcgillespie]

Glad that helped, and nicely implemented -- Your work looks very nice!

Dave

 

[sabrefencer]

Thanks again for your help.

 

[sabrefencer]

Recap of the C11 audio signal path is complete and went without a hiccup. Before doing this major surgery on an original vintage unit, now with a properly upgraded power supply, I wanted to have a performance baseline. I did a distortion test using my HP 334A (restored) with a signal generated by a Heathkit IG-72 (also restored). To keep the post brief, I am including a graph that shows the THD readings at 11 frequencies from 20Hz to 25kHz on each channel, both before the re-cap and after. Distortion went from over 1% to about 0.25% or less above 20Hz, which is very close to the original 0.2% spec. So yes, replacing those old bumble bee caps was necessary.

 

[sabrefencer]

Here are the details on the signal path recap:

Replaced C5,6,7,8,9,10,15,16,17,18,19,20,21,22,23,24 with stock capacitance value Russian Paper in Oil capacitors rated at 630v, except C5, C6, C15, C16, which are rated at 400v. Used a mix of K40P-2, K42Y-2, and K75-24 depending on applicability, availability, and size. Replaced C3 and C4 with 100uF 16v Sprague Atom TVA series electrolytic - matching spec on C22. Replaced R13, R14 with 1/2W 1M metal film resistors, as one was slightly more than 10% out of tolerance. All other resistors tested within tolerance and were left in place. No ceramic disc caps were changed. See photos.

 

[sabrefencer]

Now that the re-cap is done, I have been doing listening tests, and have found a few unwanted noises, but otherwise great sound.
1) Random static (sounds like wind in a microphone) in the Left channel only, unaffected by volume control (even at 0 setting) or any input, or control setting - it's there (intermittently). I tried different pre-out to power amp cords, even bypassed the main out jacks, different power amps and sources. I solved this (I think) by cleaning the pins on V4, of which both stages serve the left channel. First I replaced the tube - no noise, then cleaned the pins with a brass wire brush and put the existing tube back, again no noise. That was earlier today, so hoping I have this one licked. I will clean the pins on the other tubes, cuz it seems like a good idea. Sockets were cleaned along with controls early in the project.
2) A "Pop" in the left bass control, but only at one specific detent (about 11:00). This one has not been solved. Assuming the switch is intended to make-before-break, I checked that this is the case with a continuity tester. I have cleaned it several times. It is only happening on this one control on this one detent -- Any suggestions are welcome.
3) Volume control - three items. 3a) After three cleanings (CRC/Deoxit/FaderLube) I still get a bit o static on the right channel only - it is faint, but there. 3b) When I touch the knob, I sometimes get a pop, and almost always get a very faint hum - from the right channel only -- another mystery to be solved. 3c) L/R tracking - while not perfect, I connected the output to the scope and was able to use the individual L/R channel volume knobs (not balance) to set the volume tracks to be the same in the range from about 10:00 to 1:00, and then they come back together at 3:00. Difference is not huge in other positions, so I can live with this tracking, as it is not easy to hear the difference, and from what I have read, mine is probably as good as it gets for this pot at this age. Anyone have one of these babies sitting in their parts drawer?
4) Thump on power up/power down. Is this "normal", even with new PS caps? Should I consider a thermistor in the line? The thump is a few seconds after turning on power switch, as unit warms up. I cleaned the power switch several times.

Conclusions/ Next steps
I have considered replacing the RCA jacks, but that seems like a very invasive job, and I would prefer not to, unless the jacks are causing an audible issue - which so far I have not proven.
I have not yet replaced the grounding rivet on the low level input panel, and don't really know if that could be the cause of the above issues or not. Phono is pretty quiet, no hum, just white noise at high volume.
I am not ready to give-up fixing the bass control pop or the hum when touching the volume knob. Any troubleshooting or remedy suggestions welcome.

 

[sabrefencer]

I solved 2 of the 4 items above, and probably going to declare victory -- this thing sounds very nice.
#1 is still fixed, cleaned the pins on the other tubes, and she is quiet as a mouse.
#2 The pop won't go away - I tried cleaning the bass switch again, and gently adjusting the switch contact - It still pops.
#3 I went crazy with CRC QD contact cleaner, heavily flushing both L and R gangs of the volume pot, then hit it with a bit of deoxitD5. The crackle is gone, and the hum when touching the knob is gone. Yay! Just a touch of F5 fader lube, and leaving it alone. The tracking imbalance did not change from the cleaning, and it is better than good enough. With the individual L/R channel adjustments, I should not have to play with the balance control to get even levels from both channels throughout most of the range of the volume pot.
#4 Still thumping. I will make sure I turn the preamp on and let it warm up before turning on the power amps, and shutdown in the reverse order.

For now I think I am done, unless someone has any ideas about addressing the bass switch pop. For now I will not be changing the RCA jacks or the ground rivet, unless I have issues down the road.

 

[dcgillespie]

# 4 is absolutely normal. I run my C11 with my MC240 and turn them on together, and no thump -- but there wouldn't be because the MC240 needs time to warm up. If you're using the C11 with a SS power amplifier (which basically didn't exist when the C11 was in production) and you turn the C11 on either near simultaneously with the power amplifier, or certainly after the power amp, you'll get the thump. In the C11 -- the uses of instant on selenium/silicon rectifiers and R/C coupling guarantees it. With SS power amplifiers or if you warm up a vacuum tube power amplifier, simply turn the C11 on first, wait about 5 seconds, and then turn on the power amplifier.

Dave

 

[sabrefencer]

Thanks Dave. With number 4 crossed off the list, that just leaves, #2 -- "pop" on bass control. Any suggestions, or should I just live with it?

I just finished installing the C11 in my custom multisystem thru-wall rack mount. The C11 is driving restored Dynaco MkIIIs into 14" 2 way DIY JBLs. The C33 is driving an MC2155 into JBL 4311As. Head to head I'm liking the tube system more, but maybe because it is "new" to me.

 

[labjr]

Looks great!

I'll be looking at your mods since I just got in a C11 preamp that needs help. Someone kludged a few caps in the power supply. B+1,B+2 & B+3 voltages seem 10 volts or so low, but the heater supply seems a volt or so high. Not sure what the problem is yet.

Also the right channel has more gain and hum on phono, mic and tape head inputs.