LOMO Tube Mic Pre, Elevated Filaments?

So I've been messing around with one of the Soviet LOMO UP-27 tube mic preamps (schematic attached).  They have two transformer isolated inputs, mixed to one output.  The first three tubes (6J32P) are roughly equivalent to EF86, the 6N3P output is cross referenced to a 5670.  I have an original 220v power supply, but the PSU is an odd piece so I started by just using my bench supply. 

With 6.3v DC filaments and 200v DC B+ the preamp is working and gives about 75dB gain.  Distortion is not great, nor is headroom.  Now I realize that the White cathode follower output stage is not going to put out a lot of current, but this thing is into heavy distortion at +10dBu output.  Increasing input to -50 or -40dBu and lowering the gain control (R18 for channel 1) does not really seem to change things. 

From what I've read here, the B+ should be in the 250-300v range.  The reason I stopped at a B+ of 200v is because that brought the cathode of the first triode of the 6N3P up to 84v.  The max cathode/filament voltage allowed in this tube is 100v and I don't think I could get to 250v B+ without hitting 106v on the cathode.  However this brings the Grid 2 voltages of the first three tubes to 50-60v and their plates to 29-32v which seems really low.  So I tried to suss out the stock PSU without a schematic and it is kind of odd but I was able to get some voltages out of it: 300v, 35v, and 7vac all measured with no load.  That got me to thinking that the 35v might be for elevating the filaments.

Now I've read about this method but still don't think I understand how it works without burning out the heaters.  I understand that the AC is actually supplying the current for the filaments and the DC is sort of biasing it, but how does the filament "know" to draw a lot of AC current and little DC?  Why doesn't the excessive DC burn it out?

All things considered, does a B+ of 300v and an elevated filament voltage make sense for this circuit?  It seems like it needs more than 200v for sure.  Are other folks just running these at 300v and ignoring the cathode/filament rating?  Looking forward to learning!

Mike

Hi Mike,
I,m running at 250V HT and unelevated heater - 6.3Vdc on the modules i've racked for someone - had no reported probllems in over 18 months.

Go to Jakobs site gyraf.dk, scroll down left side to resources and look at the PEQ1A pultec eq circuit. They elevate the heaters using a potential divider off the HT rail.
You don't need a seprate winding for the elevation - just a potential divider across the HT rail.

The Modules supplied to me had the heater tied to 0V with a wire link underside of the module pcb.
Test that the heaters are isolated from HT 0V before trying to elevate them.

If you are using DC heaters then tie the divider to the + out of the reg circuit BUT AGAIN, make sure the heater REG circuit is floating - ie NOT tied to 0V.

I may try this with the next modules i have to rack up.

mikeyB

Oh - forgot to SAY - THE LOMO INPUT TXS ARE AWFUL - replaced with a Lundahl LL1576 i think (1 to 5 step up) and a carnhill ungapped 9600R to 600R (cant remember model number) from audiomaintenance.com. SWEET!!!

mikeyB said:

If you are using DC heaters then tie the divider to the + out of the reg circuit BUT AGAIN, make sure the heater REG circuit is floating - ie NOT tied to 0V.

Click to expand...

Hi Mikey,

Thanks for the replies!  But are you saying here that you can have DC filaments AND add a DC offset???  That just makes no sense to me, but maybe I'm not understanding correctly.

As for the transformers, it seems like the stock unit has a bit of a "smile" frequency response which doesn't really bother me as a flavor sort of thing.  Does the Lundahl/Carnhill flatten it out or just sound better?  Also, do you lose much output level with the step down?  I was considering a 1:1 output with low DCR.

Mike

Follow the moneyXXX current path.

The heater is connected to ONE side of the 84V DC supply. And NOT the other side. It has NO way to flow the whole 84V.

I give you a 3-volt flashlight. Then stand you on top of a 100,000V Tesla Coil. The flashlight works fine, because it has no real way to know it is 100,000V above the other end of the Tesla Coil (ground).

(YOU may know there's 100KV under your feet, because even slight leakage can lift hair off your head and cause corona-crackle off your fingertips.)

If 6N3P is a 5670 which is a 2C51, that's a lightweight line driver. It has a nice low small-signal output impedance, but the large-signal impedance is many K ohms. There's also 6k8 in its feed. (R29 could perhaps be 270r, but that's a minor thing.) The 4uFd output cap suggests >2K nominal loading. If you have trouble driving 10K, then post the WCF midpoint voltage.

The more I look at it, it doesn't seem to be any kind of "studio grade", nor even good PA stuff. C10 R23 suggest there's a heavy bass-cut they are trying to hide, which may be the input transformers?

Elevating heaters is quite easy. Normally you would have one side of the heaters connected to 0V of the HT supply. To elevate the heaters, instead of connecting one side to 0V you connect it to a decoupled pot divider across the HT.

Cheers

Ian

OK, so the elevated heater has no ground reference in common with the B+.  Still seems counter-intuitive but I can get my head around it ;-) 

@PRR - yes, it was suggested to me by a knowledgeable friend that R29 could be at least cut in half.

A bit more digging show the specs for the UP-27 to be 40-14000 Hz, sensitivity 0.2 mV, output 0.775 V.  So it's meant to have 0dB output and probably goes to a buss or other amplification stages.

Perhaps I'll try it out with DC filaments and 250v B+ and see how it specs. 

Thanks for the insight.

Mike

mjrippe said:

OK, so the elevated heater has no ground reference in common with the B+.  Still seems counter-intuitive but I can get my head around it ;-) 

Click to expand...

It does have a ground reference at audio frequencies. That's why it's decoupled to the B+ ground.

Cheers

Ian

mjrippe said:

mikeyB said:

If you are using DC heaters then tie the divider to the + out of the reg circuit BUT AGAIN, make sure the heater REG circuit is floating - ie NOT tied to 0V.

Click to expand...

Hi Mikey,

Thanks for the replies!  But are you saying here that you can have DC filaments AND add a DC offset???  That just makes no sense to me, but maybe I'm not understanding correctly.

As for the transformers, it seems like the stock unit has a bit of a "smile" frequency response which doesn't really bother me as a flavor sort of thing.  Does the Lundahl/Carnhill flatten it out or just sound better?  Also, do you lose much output level with the step down?  I was considering a 1:1 output with low DCR.

Mike

Click to expand...

The output valve won't drive a low impedance 1:1,  and 75dB is some gain. I provided a switch at the output so that could have unbalanced at full output or with the tx drops about 12dB The lundhal at the front gives much better response and it is time to tinker with the various eq bits of the circuit to improve upon the original.

PRR said:

The more I look at it, it doesn't seem to be any kind of "studio grade", nor even good PA stuff. C10 R23 suggest there's a heavy bass-cut they are trying to hide, which may be the input transformers?

Click to expand...

Yeah - first module i received had been wired unbal at the input (haha - cheeky ebay seller had the transformers taken out of the cans and the cans welded up - hence the unbalanced wiring!!! - empty ali can makes a sweet little cowbell though!!)

Actually sounded ok - the front end bit of filtering was taken out too!
Fitted the lundahl to enable condensor mic use.

PRR - C10 R23 are a bass lift then? - how much is it doing and at what frequencies?

Tried it this morning with 250v B+ and the output and distortion were both MUCH better.  Talk a bout a bass lift though, I measured +7dB at 100Hz compared to 1kHz!  Changes a bit depending on source and load impedance, but always a low end boost.

> +7dB at 100Hz compared to 1kHz!

Change C10 to 0.02uFd. That should flatten the last stage, let you find any other quirks.

PRR said:

> +7dB at 100Hz compared to 1kHz!

Change C10 to 0.02uFd. That should flatten the last stage, let you find any other quirks.

Click to expand...

As always, thanks PRR - yes - with the lundahl on the front it can seem a bit bass heavy!
I am just racking another pair so i will experiment with this. They do actually sound good once the transformers are sorted out.

Sorry to resurrect an old thread, but it conveniently already has the schematic in it. I'm racking a few of these for clients and wanted to know if anyone has made any more headway with these.

On a few of them, I'll be bypassing the mixing tube and adding a second WCF stage to make them true 2-channel modules. As most of the filtering is happening around the mixer, I'm hoping that will clean up some of the anomalies.  For those that have gotten these working, how effective is the "correction" control? What frequencies does it seem to alter? Does switching out the input transformer change the response of this control?

I'll try to take some pictures and note observations as I go.

any news  ?

OneRoomStudios, do you have a result with bypassing the mixing tube ?

Hi.
I know that fhis is very old topic but can anyone tell me how fo control the volume of this pre?I don' have the original fader so i'm geftong very high level thos way...Thanks!o

OneRoomStudios said:

Sorry to resurrect an old thread, but it conveniently already has the schematic in it. I'm racking a few of these for clients and wanted to know if anyone has made any more headway with these.

On a few of them, I'll be bypassing the mixing tube and adding a second WCF stage to make them true 2-channel modules. As most of the filtering is happening around the mixer, I'm hoping that will clean up some of the anomalies.  For those that have gotten these working, how effective is the "correction" control? What frequencies does it seem to alter? Does switching out the input transformer change the response of this control?

I'll try to take some pictures and note observations as I go.

Click to expand...

G'day folks - I'm also super keen to hear what seems to be 'the way to go' with this module.

I have a rack here, on loan from a mate here downunder - racked up by Rob Squier some years back..

It seems to me to exhibit the 'low cut' response that some have mentioned, so I was interested to find there that others have ended up with a bass lift!

Anybody got photo's of wiring or notes on component configurations that worked out nicely for you?

Cheers!
Evan

Personally I would be sorely tempted to change the otuput stage from a White follower to an SRPP and at the same time change the tube to a 6H23 (Russian equivalent to 6dJ8/ECC88). This will significantly improve the drive capability and the extra loop gain should educe distortion. Add a chunky Carnhill 9K6:600 output transformer, increase the 6Uf output cap to 10uF and it should work a lot better.

Cheers

ian

..but note that the 6N23P tube draws almost double the heater current as the ECC88, so make sure that you can supply that..

/Jakob E.

gyraf said:

..but note that the 6N23P tube draws almost double the heater current as the ECC88, so make sure that you can supply that..

/Jakob E.

Click to expand...

Hi Jakob. I am not sure which tube data you have been looking at but both the 6N23 and ECC88  tubes consume 365mA at 6.3 volts.

Cheers

Ian