Skip to main content

Can this be right?

It's just been pointed out to us that our Pultec MEQ-5 drops the signal level by almost 4 dB, which we've confirmed in our lab with a signal generator / analyser.

After changing ALL the valves the relative o/p level is now -3.5dB but I don't know if this particular unit is supposed to be "zero in, zero out" or if this signal drop is normal.

Any idea anyone?

Phil
Techie, Olympic Studios

Comments

Boswell Mon, 06/11/2007 - 03:01

The passive EQ section of a Pultec has an insertion loss of 16dB. In the MEQ-5, this is supposed to be made up to 0dB by the gain in the output stages. However, it is an open-loop design, and hence not accurately gain-controlled. Check the anode and cathode-cathode resistors of the first triode stages and the cathode by-pass capacitors of the output triodes.

Thomas W. Bethel Mon, 06/11/2007 - 04:47

Also how are you running the unit? Balanced in and out, unbalanced in and out or a combination? When was the last time the tubes were checked and were they checked for gain (mu on some tube testers). Do you have the right tubes installed. A 12AU7 looks the same as a 12AX7 and will work the same in a circuit but has less gain.

Here is a simple way to check on the gain of your tube(s) http://www.rru.com/~meo/Guitar/Tubes/t3.pl

anonymous Mon, 06/11/2007 - 05:04

Thomas W. Bethel wrote: Also how are you running the unit? Balanced in and out, unbalanced in and out or a combination? When was the last time the tubes were checked and were they checked for gain (mu on some tube testers). Do you have the right tubes installed. A 12AU7 looks the same as a 12AX7 and will work the same in a circuit but has less gain.

Here is a simple way to check on the gain of your tube(s) http://www.rru.com/~meo/Guitar/Tubes/t3.pl

Balanced in and out.

The tubes that are in there now are all brand new - just arrived this morning, and are all direct replacements for the ones that were in there before I changed them (I'm assuming they're the correct ones and that someone hasn't put the wrong valves in there previously).

They are: a 6X4, an ECC82(12AU7) and an EL90(6AQ5).

anonymous Mon, 06/11/2007 - 05:19

Boswell wrote: The passive EQ section of a Pultec has an insertion loss of 16dB. In the MEQ-5, this is supposed to be made up to 0dB by the gain in the output stages. However, it is an open-loop design, and hence not accurately gain-controlled. Check the anode and cathode-cathode resistors of the first triode stages and the cathode by-pass capacitors of the output triodes.

Yeah, it's time to open it up, isn't it :-)

Thanks for your help, both of you guys.

Boswell Mon, 06/11/2007 - 05:27

I've actually had a look at a UA Pultec circuit, and I had mis-remembered the output configuration on those types. The output stages are a closed-loop design via a feedback winding on the output transformer that differentially drives the first cathodes. So if you are getting a no-load overall gain of around -4dB it means either that the passive input stages (including the input and interstage transformers) are attenuating more than 16dB, or the open-loop gain of the output stages is insufficient to make up the 16dB when the transformer feedback is applied. The fact that you see a gain difference when you try different valves would point to the latter.

Tom's link to the gain check site is interesting, but not easy to apply to circuits that have transformer-coupled feedback.

BTW, the gear list at Olympic Studios is impressive!

Kev Mon, 06/11/2007 - 14:27

what lab with a signal generator / analyser are you using ?

have you confirmed that it will hold output level ... when terminated with Transformer input stages like the MEQ-5

beware the diferences between bridging input and their relatively high impedance
and
old styled 600 ohm ... and lower ... transformered inputs

I think people around here already know I like to use passive VU meters at both input and output, to confirm these sorts of measurements

anonymous Wed, 06/13/2007 - 04:48

Kev wrote: what lab with a signal generator / analyser are you using ?

have you confirmed that it will hold output level ... when terminated with Transformer input stages like the MEQ-5

beware the diferences between bridging input and their relatively high impedance
and
old styled 600 ohm ... and lower ... transformered inputs

I think people around here already know I like to use passive VU meters at both input and output, to confirm these sorts of measurements

It's a Neutrik A1 "Audio test and service system"

anonymous Wed, 06/13/2007 - 05:48

OK - I've stumbled across something...

while testing these resistors, I was unsoldering one leg of each component in turn for testing, then resoldering afterwards.

after all the obvious resistors, I unsoldered one leg of what I had thought was a resistor before unsoldering, which turned out to be an inductor, however when that one leg was free, the whole thing dropped out of the circuit, so it must've had a dry joint on the other leg. (promising sign).

The problem is, because I was concentrating on the leg I was unsoldering, I hadn't taken any notice of where the other leg was attached.

Does anyone have a schematic that they can either let me have a copy of, or could anyone that has a schematic just inform me of which terminals this inductor should be soldered to?

anonymous Wed, 06/13/2007 - 07:24

Right - I just powered it up without the inductor in circuit and I'm getting the same result... 3.3dB down... which means that is very likely to've been the problem.

Can anyone let me know where it's supposed to go? (y)

P.S. for the record, the rest of Olympic's maintenance guys are much better than me at this sort of thing - I'm the "computer guy" :-P

Boswell Thu, 06/14/2007 - 02:03

Jakob Erland's site Gyraf Audio http://gyraf.dk/ has some schematics from which you can piece together most of the circuit of the MEQ-5. Go to the Resources link under Technical Documents and scroll down to Equalizers. The MEQ-5 EQ section is there, and the output stages can be taken from the PE1A schematic.

If you are experiencing gain loss in bypass mode, the problem is unlikely to be due to poor signal inductor connections. However, my memory of those Pultecs is that all the signal inductors are tapped and so are not 2-leg devices.