MEQ-5 questions

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Deepdark

Well-known member
Joined
May 19, 2013
Messages
1,321
Location
Quebec, Canada
Hi guys

Some questions aout the amp section. I plan using it on a eqp1a, but had some questions.

1. The first stage (12ax7) looks to be some sort of phase splitter stage, according to Morgan jones  valve amplifier book (4th edition). Page 459, figure 6.19. Never seen it before. Any things I should take in consideration/take care of to avoid any trouble?

2. The output transformer is fed by the cathode of the 6aq5. At the node before the 18K - 5W resistor and the primary high, it shows +160v!!! Is it a typo? straight into the transformer??? looking at the 18k resistor, if we need to drop 160v throughit, it gives us about 8.8ma, about 1.5W, so our 5W looks logical.

3. 1k resistor between heater ground and B+, heater elevation I guess? essential in this topology to avoid hum?

Thanks guys!
 
Deepdark said:
1. The first stage (12ax7) looks to be some sort of phase splitter stage, according to Morgan jones  valve amplifier book (4th edition). Page 459, figure 6.19. Never seen it before. Any things I should take in consideration/take care of to avoid any trouble?
It looks like two inverting sections and a follower.

Deepdark said:
2. The output transformer is fed by the cathode of the 6aq5. At the node before the 18K - 5W resistor and the primary high, it shows +160v!!! Is it a typo? straight into the transformer??? looking at the 18k resistor, if we need to drop 160v throughit, it gives us about 8.8ma, about 1.5W, so our 5W looks logical.
There is a cap on the other end so there's no DC.
 
Thanks. So even if the cap is at the primary low, it will block any dc, even at the primary top? Wouldn't it be better off to put the cap before the primary top, just like an la2a, in exemple?
 
Deepdark said:
Thanks. So even if the cap is at the primary low, it will block any dc, even at the primary top? Wouldn't it be better off to put the cap before the primary top, just like an la2a, in exemple?
I would just implement the circuit exactly as it was intended according to Pultec. Otherwise what's the objective really? If you want to make it "better" then one could argue that you should replace the entire circuit with a few op amps.
 
> sort of phase splitter stage

Not nearly. The popular Cathodyne splitter has =equal= resistors top and bottom. What do we have here? Over 1000:1 difference.

I agree it is drawn funny. Re-cast it in conventional shape.

It is two voltage gain stages and a cathode follower. A tap on the CF load resistor goes to the 1st stage cathode to give it bias and NFB. Pencil-check suggests forward gain around 1,300, but 18,000:150 gives 121. There's >10X more gain than needed, so the NFB is effective, and final gain probably 8% low or 117.

The ohohoh... P-G cap in the 2nd stage is 10pFd. We can throw 30+pFd at a 12AX7 and not change audio-band gain much. So this is to control supersonic gain. We have two gain stages and a wonky transformer load, we just might get the three phase-poles needed to oscillate, they apparently thought the 10pFd was necessary and sufficient.

I have no idea why there is 1K in the 6AQ5 plate lead. It isn't filtering. The 6AQ5 triode will be over 2K plate resistance, I don't see why they artificially increased that. It isn't "protection" (as we might see in a transistor EF) because the 6AQ5 is not going to gush current.

Tying "high voltage" to an output transformer is perfectly safe and common; that's how all Power Output stages are rigged. The HS-50 was made for this. There's no special reason to put the coupling cap top or bottom; often wiring convenience decides this.

I agree: do it like they did it in Englewood, or leave the past behind and throw in an opamp. Voltage gain can be 30 or less since you don't want a 10K:600 output transformer.
 
Thanks guys! I agree there is way too much gain in this thing. Would it be a good idea to drop a 100k pot before the first stage to adjust the gain? Just like an la2a? An eqp filter stage has about 20db loss, no need for much gain! Maybe droping a 12ay7 instead of the ax7 help taming the gain, Too.

I already planned an opamp version, and an 600:600 input and output transformer.
 
The gain make up amp is just a variation on the MB-1 mic pre without the variable gain.

133771d1251343261-help-mb1-diy-mb1-schem-


To alter the gain you could just reinstate a version of that gain control circuit.

Cheers

ian
 
Hi Ian. You're right, the first stage is quiet similar. The input transformer for that circuit would be around 1:4 or 1:5 i guess. Anyway, the eqp filter stage is 600ohm or so.  Output transformer ratio around 4:1 or 5:1 would do the job.  Looking at the putput transformer connection, they add the dc blocking cap at the primary high. Why wouldn't they do this with the meq-5?

Looking at the psu, i will replace the 6x4 by a full wave rectifier arrangement, and add some filtering stafe (probably work it out for a 5 RC filtering stages. Heater looks to be elevated, right?

Not quiet sure to get the gain selection on the mb-1. Could you explain it please? The 160k/1k5 resistor form a voltage diviser, so making the 1k5 variable would set our gain?
 
I think the 160K and 82K plus the cap are just so the can wire the transformer in direct and get the correct secondary load across it. You don't need that. Just connect the junction of the 1K5 and 120 ohm to the first triode grid via a 470K resistor. This means the grid is now quite a few volts dc above ground so you need to add a 100nF  capacitor in series with the input to couple it the the EQ output. The gain is sett essentially by the ratio of the 17K5 to the 1K5 so it is nominally 21dB. In the original they vary thee 1K5 to vary the gain but this does alter the operating point of the first triode as you change gain. Instead you can  add a 200uF in series with a variable resistor across the 1K5 to vary the gain.

The only problem with this design is the gain cannot be less than. I use the same principle in my Eurochannel mic pre but in effect I change the 17K5 and 1K5 to 10K and 10K which makes the nominal gain just 6dB. A 200uF cap and series resistor from the first triode cathode to 0V sets the desired gain. The downside of this is that the first stage cathode is now raised to about one quarter of the HT supply so you need to reduce its plate resistor to maintain the right idle current.

Cheers

Ian
 
ruffrecords said:
I think the 160K and 82K plus the cap are just so the can wire the transformer in direct and get the correct secondary load across it. You don't need that. Just connect the junction of the 1K5 and 120 ohm to the first triode grid via a 470K resistor. This means the grid is now quite a few volts dc above ground so you need to add a 100nF  capacitor in series with the input to couple it the the EQ output. The gain is sett essentially by the ratio of the 17K5 to the 1K5 so it is nominally 21dB. In the original they vary thee 1K5 to vary the gain but this does alter the operating point of the first triode as you change gain. Instead you can  add a 200uF in series with a variable resistor across the 1K5 to vary the gain.

The only problem with this design is the gain cannot be less than. I use the same principle in my Eurochannel mic pre but in effect I change the 17K5 and 1K5 to 10K and 10K which makes the nominal gain just 6dB. A 200uF cap and series resistor from the first triode cathode to 0V sets the desired gain. The downside of this is that the first stage cathode is now raised to about one quarter of the HT supply so you need to reduce its plate resistor to maintain the right idle current.

Cheers

Ian

Thanks Ian. I guess you mean something like this?
 
here is a pcb of it, without the psu for now. I'll make the psu on another board. I tried to keep the leed as short and straight as possible. 
 
Deepdark said:
Thanks Ian. I guess you mean something like this?

Clse but there are a couple of bits missing (see attached pic).

The top shows the simple mode to reduce gain. The 18K gets split into 16K5 and 1K5 so the CF current stays the same. The 12AX7 cathode gets raised to about 15V so you need the 100nF at the input. The 200uF plus resistor allow you to increase the gain above the nominal 21dB

The bottom shows a version arranged for 6dB nominal gain with the 18K split into two 9K. The cathode now rises to 90V so yopu need to reduce the plate resistor to raise the plate voltage. 150K is shown but I do not think this is low enough; it probably needs to be 75K to maintain the same operating point but this will reduce the first stage gain. You might also need to adjust the bias for slightly lower current through the 12AX7 by increasing the cathode bia resistor.

Cheers

Ian

 
that 10 pf could be done with chassis wires,

that 1 K resistor is a mystery, could it be there to protect the transformer primary?

no, even if the cap shorts or the tester guy gets his needle nose across it, you still have 300 volts into 2.2K (1.2 K pri DCR) which equals 136 ma into #46, which is rated at 2.5 ma,

maybe it is there to tweak the bias voltage, is there grid current flowing in the positive half cycle like a traditional follower? don't know, probably need a load line or something,

is there a phase flip in this box? Pac Bell don't care but Capitol Studios might, looks like someone was trying to figure that out with sq wave symbols,

it is a great sounding circuit, just as good or better than the EQP-1A

be a shame to switch to solid state with such an easy wiring job,

better to build a Lang PEQ or use a Melcore DOA than some Burr Brown sizzle chips,  :D

OT - here is a very good article on the DC coupled Cathode Follower, do not know why somebody has not implemented a pot in the cathode circuit to adjust compression for intentional distortion in a guitar amp,

http://www.valvewizard.co.uk/dccf.html

 
Back
Top