Collins 6q-1 mic preamp pics and schemo

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Yes, and I can't think of any other examples of such low filter capacitance in the post-1932 era.  There are a total of 2 mfd filtering for an entire channel.  Normal is 10 mfd or more per stage, this is a design theory from an earlier time, still used as tried-and-true, and more bulletproof for lack of electrolytics.  The 50 mfd bypass cap on the 6c4, earliest version, is the only thing you ever find bad or previously replaced in one of these.  1 mfd is low enough to not pass bottom frequencies to B-, thus boosting the lows a bit. 

Note the extra large interstage coupling cap.  The output stage coupling cap is standard here. 

All the caps here are hermetically sealed, metal cased, paper in oil 'bathtub' types. 

Note the extremely low plate resistances; matching condition or less, rather than 2x or more.  Lassoharp pointed this out recently, and my initial thoughts were:

6Q headroom is same for each stage, based on field observation.  Rare; usually the front end is much lower, here the front end is higher and the output is lower than normal.  PRR pointed out somewhere that this is an old trick to cancel distortion.
6AQ6 stage at 47K about 1/5 what you normally see.  Less than matching condition. 
Not going for gain, so much as current to drive mixing bus?  70 dB in the 6N program amp.  Output headroom is relatively low in 6Q. 
Early 30's power tube POV is that matching condition is lowest distortion, and double plate resistance load is best balance of voltage gain versus distortion.  PRR disagrees below, and I can't argue other than to reference some articles from the era. 

6Q already weird using a high mu 7 pin small bottle front end in 1947.    I go on about hating 12AX7's, and here we have half of one.  And 6AQ6 sounds good here.  But, not as usually implemented. 

On THD, whatever it is, if you want it, it's gorgeous here.  These are the quintessential tube overdrive pieces, IMO.  You need to replicate early Little Richard records, this is your easiest ticket.  I hesitate to use words like creamy, but some electric guitars through one of these get 200% better than they actually are, immediately.  Distortion content has a clearly audible 20 dB curve when you put an interstage volume control in place.  I have never found that in any other circuit. 


 
emrr said:
I have never found that in any other circuit.

Me neither, not that I would have access to many classics anyway. But I followed some of these fourties-fifties broadcast preamp design ideas and gain staging and stole some ideas for a bit more flexible version of my own. Then I finally tuned the plate resistances and biasing by ear, and it now looks awfully similar to this collins.

emrr said:
Rare; usually the front end is much lower, here the front end is higher and the output is lower than normal.

This might be a major reason it sounds as good distorted as it does. I had to do the same because I only had a 2:1 ratio output transformer around. First stage (two stages in mine) quite linear, some local feedback and high plate resistor (and gain of course). Output stage (ECC99 on mine) has "abnormally" low plate resistor so it would gracefully drive the transformer, don't care about gain here. It tries to drive the output with as little distortion as possible and fails, but gracefully. Sounds damn good. Mostly 2nd harmonic and a bit less than 1% overall THD when not clearly overdriven/clipped. I have not tried a 6C4 tube anywhere, but I suspect it's a similar behaviour with collins' "low" 10k plate resistor.

That 680k feedback, I think it might be there mostly for better low frequency impedance, since it won't do much anything to overall THD. Take it out and there might be a slight low frequency cut, depending on the output transformer of course. For a low ratio transformer it will have a clear effect when the output stage is struggling to drive the output.
 
The plot is a unit lacking the FB resistance, but having the 6C4 cathode bypass.  The later lack of cathode bypass on the 6C4 improves bass, and my gut said the slight FB was there purely to even out gain variation between units.  

My hack 20 dB FB pad seen in the schematic gives the same response on the bottom, and flattens the top bump completely. 
 
emrr said:
The later lack of cathode bypass on the 6C4 improves bass

I wonder how. Doesn't the lack of bypass cap also increase the output impedance, and especially low frequency driving capability?

But then again it's also local feedback... I will have to experiment with this at some point.
 
50 mfd here seems to have a measurable effect.  The one time I experimented years ago, it took 470 mfd to get as flat as none, in this amp. 
 
Are R105 & 113 220 Ohms or 220K please anybody.....I know the resistors marked M mean K....thanks
 
Thanks Doug, I tried both seems to sound the same but will stick with the 220R.....I just soldered up a P2P version of this  circuit using Sowter transformers, I seems to sound a bit fuzzy not in a bad way I seem to remember you saying something about these distort in a good way.......tried B+120V upwards distorts less at higher B+.....bit more experimenting to do now....thanks for info.
 
Aha when remeasured my B+ under load I found my B+ very low.......have since tried 100V and or 190V    100V sounds fuzzy but interesting 190V is cleaner  I guess this is normal......Is 190V too much tho will I hurt my tube/tubes.....if thats the case il have to make a little potential divider.....

I have messed with 680K feedback resistor on or off doesnt seem to make much difference to my ears will prob take it off....

Can I put a 47K pot in place of R102 C103 to wiper ?

I have a big fat oil .25U I might try for C103 wonder if that will sound good.....the thing so far sounds great I love it.......need to make it a case of some sort now.....
 
> Early 30's power tube POV is that matching condition is lowest distortion, and double plate resistance load is best balance of voltage gain versus distortion.

No.

For "transformer coupled" (DC in transformer):

If drive is limited, Z-match is maximum power output. It is also nearly the highest THD (a hair better than a low-low-Z load).

If excess drive is available, 2X will make more output with lower THD and without high supply voltage.

If drive and supply are ample, 3X-5X can be rigged for higher efficiency, more power, and lower THD.

To get well past 2X load you are tempted to go to a low-Mu tube. However this needs more drive, so you go around the design-circle again.

Maximum efficiency is 50% with perfect devices, 20%-40% with reasonable real devices.

For Resistance-coupled (as used here):

The MAXimum power efficiency is 4%. So if non-negligible power is required in load there is NO "good answer". Design usually is guided by trying to get enough load power, and let the other parameters fall where they may.

With only 140V 8mA supply, they were not being greedy. 6mA in final times 140V is 0.84W consumed. Very-best output with an ideal device is 0.033W or +15dBm. 140V at 6mA is 17K total DC load. 6C4 is a readily available 7K triode. You are going to pick the plate resistor near 10K, and the load somewhat higher, 20K is reasonable. Actually for 20K load and 7K triode the optimum DC resistor is 7K. Going 10K lowers THD but reduces max output more. Going to 15K6 DC for bass-bump is even less efficient.

Obviously they had a rather low Max Output spec (not given on that page). Perhaps +8 or +10dBm, 0.010 Watts. (Probably marked nominal -10dBm or -18dBm on the console plan.) They had a 3.5W bottle and could afford 0.9W of DC power. Required efficiency is 0.01W/0.84W= 1.2%. This is short of the maximum 4%, so they could use a not-MAXimized design, but not too far off the maximum.
 
PRR said:
> Early 30's power tube POV is that matching condition is lowest distortion, and double plate resistance load is best balance of voltage gain versus distortion.

No.

Quoting a series of 1930's articles and texts there, broad stroke.  There is an article in Electronics, maybe 1932, which gives many charts and graphs showing how power tube distortion is lowest with a matched load.  Thanks for the breakdown. 

I recall getting about 0 dBm clean out of these at stock voltage.  I usually run them up to 200V B+.  No reason to not go higher, if willing to consider higher rated resistances.    I remember current doubling at about 220V B+. 
 
Hi Doug !

First of all, thanks for all the information you have provided about old tube gear throughout the forum ! I really enjoy reading your posts with this nice mix of technical explaination, historical background and passion about recording.
I got particularly interested in the 6Q-1, probably because you frequently mention it as one of the best kept secret regarding 'dirty' preamps.

emrr said:
6Q already weird using a high mu 7 pin small bottle front end in 1947.    I go on about hating 12AX7's, and here we have half of one.  And 6AQ6 sounds good here.  But, not as usually implemented.  
As this preamp is difficult to find (especially in France) and I see this preamp as 'just' a two-stage preamp with an uncommon biasing point for the 6AQ6, I was thinking about making an updated version that anyone could build with film caps, half a 12AX7 (instead of the 6AQ6) and half a 12AU7 (instead of the 6C4). I would also like to prove that one can make a nice sounding dirty preamp with a 12AX7 ( my personal take of 'I want to believe' crusade :) )
I ordered most components for it and one week later, I found an original 6Q-1 on eBay for $325 (not bad !) that I purchased. More than the pride of owning this piece of gear that I have been seeking for a long time on eBay, that will also be a nice comparison point for this updated clone.

I have two questions for you and other forum members who are welcomed to answer, of course  :)  :
- I still haven't powered up the old rusty 6Q-1 and I am concerned about the old paper oil caps. With your restoration experience, did you notice that these kind of (expensive) caps had to be replaced on the ones you've seen ? I don't have much experience with them and I have read many stories of this kind of caps turning leaky over time...
- though it will not stop me from experimenting and trying it, what do you think about my idea of an unfancy 12AX7-12AU7 version of the 6Q-1 ? Any advice/things to be aware of ?

Thanks in advance !
 
This one escaped me, sorry.    You got a great deal on that 6Q-1.  Thordarson or Freed transformers?  There's apparently also a square can input transformer that I've never seen in person. 

If the caps measure fine, they are fine.  I rarely see bathtub caps that have failed, or leak DC.  If they leak oil, you may have a problem.  Note if yours has the 25 mfd cathode bypass caps, they are electrolytic and should be replaced.  I like 100 mfd here, that's usually what a good old 25 will actually measure. 

Tubes; the 6AQ6 is something like $1-2.  Sometimes they are on blow-out sale for $0.50.  Can't find a 12AX7 that can compete with that, nor can you buy a giant pile to select for low noise like you can with the 6AQ6.  Try it anyway. 
 
Thanks for your reply, Doug !
My 6Q-1 has the Thordarson transformers except for one of the output transformers which was replaced with a Freed one (677-0109-00). It does not have the cathode bypass caps on the output tubes.

I have not found the time to work on this 6Q-1 restoration project nor on the 12AX7/12AU7 version. I will report back with sound samples to compare when it is done.
 

Latest posts

Back
Top