EQP1A with cascode amp circuit

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 all

I'm looking to make an EQP1A but with a cascode amp stage. The goal is to get rid of the interstage transformer and the feedback circuit from the output transformer. All of this in a way to save some $, simplifying the circuit and made it more diy friendly to use with a variety of transformer. I know Gyraf made a SRPP design. But since I'm in a Learning stage in tube world, why not trying to designing something of my own :)

first thing first, I would like to know if it's a good candidate for such an task? Assuming the filtre section will eat up around 20db, and around what, 6db per transformer??, I know the cascode can deliver really high gain.  I consider 12AX7 in my design. I would like to know if it looks ok. Any bit of informations is welcom, since I'm in a Learning stage, I'd like to learn as much as I can :)

Looking at the transformer, I don't know how to select the right one (impedance Wise). Looking at the anode load resistor (100K) should I consider it as the circuit impedance?
 

Attachments

  • 201501131639.pdf
    469.2 KB · Views: 57
Will an input transformer rated 1:10 is preferable?s there a general roule of thumb for selectionning the right transformer?
 
Deepdark said:
I'm looking to make an EQP1A but with a cascode amp stage. The goal is to get rid of the interstage transformer and the feedback circuit from the output transformer.
just asking: what does the complexity of cascode have to do with your goal of no interstage or trick output transformer ?
a time tested output stage that can use  variety of iron is the Pultec EQH circuit,
cheap and effective. 
 
The only goal was to get rid of the interstage and the feedback winding circuit, just like the poor mans eqp or the gpultec. I just wanted to see if a cascode gain stage could be usefull and a good candidate for such a task (gain stage) I know the circuit of the eqp isn't that hard, I've made 2 of them. But There is not a lot of company who made the spécial output transformer and the feedback circuit can be tricky to deal with with, for exemple, cinemag cm-9600t.
 
> I know the cascode can deliver really high gain.

Into a VERY high load impedance.

Basics: cascode has super high output impedance. Transformers have medium/low impedance and like to be driven from low impedance. The optimum specs for your output transformer are unobtainable at any reasonable cost, maybe not at any cost depending what you want for bandwidth and clean headroom.

12AX7 is an exceptionally low-current tube and almost never a good choice for driving transformers.

*OR*: Assume the long line of audio dead-men were not fools. If a 12AX7 cascode were a good plan for driving an audio transformer, someone wudda done it. I don't think you will find any such thing. Maybe not even in the lame world that is the modern internet.

A Line Output is generally a small Power Amplifier, not a goosed Voltage Amplifier. Driving transformers and long lines and multiple loads is Real Work.

> Assuming the filtre section will eat up around 20db, and around what, 6db per transformer??

10K:600 OT is a 4:1 voltage ratio, -12dB. IF the filter is -20dB, you need 32dB of voltage gain, or about 1:40.

To drive a 10K load well you need a tube of lower plate resistance. Straight (no cascode) 12AX7 is 60K plate impedance, way out of sight. 12AU7 is 6K, a better bet. 12AU7 working high current into a 22K plate load, with a 10K cap-coupled load, can give gain of 10. The other half using standard R-C amp tables can give gain of 16. Total gain 160. At this point you really should bring NFB around from V2 plate to V1 cathode to turn your 4:1 excess gain into lower THD and better transformer drive. Since 15v peak output on the 600 side is 60V peak 120Vp-p at the plate, and the plate is heavily loaded, supply voltage must be much-much more than 120V, perhaps 300V to not have gross THD at high level.

> Looking at the anode load resistor (100K) should I consider it as the circuit impedance?

Straight (no cascode) 12AX7 will be about 60K of internal plate resistance, 100K external resistance, call it 38K effective source resistance (a poor fit for 10K transformer).

Cascoding increases the upper triode's internal resistance by a factor of Mu, 100 in this case. So 60K becomes 6,000K or 6 MEGs. In parallel with the 100K external resistor is 100K for practical purpose (2% error), and a horrible fit for any available full-range audio transformer.

Nit-picks: I have real doubt about how you have set-up the upper grid. If the bypassing were adequate, Vgk is zero, the upper tube will be drawing grid-current on any substantial swing. Even in TV Tuners (the cascode's best fit) it is often better to nail the upper grid to a fixed voltage (voltage divider). Also to bypass it across the audio band; 100pFd is hardly a serious bypass, it may take 0.1uFd to really hold the grid *constant* enough to give plausible cascode action.

For a 1-bottle design with generous power supply you might work with a TV V-sweep dual dissimilar triode. The small side has Mu of 20 to 60 and can give gain of 15-40 into a hi-Z load like a grid. The fat side can be run at large current (10mA-20mA) with low internal and external resistance, to smack the output transformer. I recall 13EM7 is one, but there are many others, some with 6V heaters if you wish.
 
I think a cascode is not a good choice, for several reasons.

1. As PRR pointed out, it has a high output impedance and a very low standing current, both of which mean it is unsutable for driving anything other than a high impedance load
2. The cascode is basically a common cathode with a grounded grid strapped on top. Unfortunately, the grounded grid has a low input impedance which means the common cathode stage distorts rather a lot. If you want signal level high enough to drive a step down transformer I suspect the distortion will be intolerable.

Gyraf's SRPP using an ECC88 is a very good one tube solution.  The SRPP is a push pull stage so it can output twice the current that a single ended stage running at the same idle current can. The ECC88 has a low plate resistance so it can drive a transformer and load at relatively low distortion and it has just about enough gain to make up for the EQ and a 2: 1 transformer. It's only downside is that the SRPP does produce more distortion than some other topologies but this only becomes significant at very high output levels.

I like PRRs idea of using a different type of tube altogether. How about something like the ECL82 which has a triode first stage followed by and output pentode. You could make a nice single ended amplifier with enough gain and drive capability and possibly even include some NFB without needing to use an extra transformer winding.

Cheers

Ian
 
Wow Thanks guys for all those information. I'll look at the srpp circuit as suggested. Is there another common circuit that could be a good candidate? Looking st the suggestion of other tube type, you strap it in a srpp or there is a specail circuit for those tube?
 
Dear Deep Dark,

You may want to have a look at the amplifier stage for a Pultec MEQ5. This circuit uses no interstage transformer. The circuit is a cascade amplifier using an ECC83 and then a 6AQ5 as a cathode follower is also uses internal feedback which you could adjust on test to get your unity gain through the unit. The output transformer used was a Triad HS-50 which I recall as being a 15k:600ohm.

I am not sure what would be a good equivalent modern transformer for the Triad HS-50, anyone got any ideas?

Also, looking at the Pultec MB-1 cuircuit, it is pretty much identical. In it, the have ommited the 10pF capacitor from the plate to the grid of the second stage of the ECC83, anyone know what purpose of that capacitor was. They also replace the 6AQ5 with another ECC83 doubled up, anyone know why they did that also?

 
Thanks man. The circuit in the MEQ-5 looks fine. I'm sure plenty of manufacturer made 15k:600 transformer. I'm pretty sure Cinemag, Jensen, Lundahl, OEP, Edcor, and so on, have somethig similar. It does not looks like a typical cascode to me, though.

Is there any books, any good source to find circuit and explanation of them, to help designing and understanding them? I searched on the web but only find some bit of information here and there. Thanks :)
 
Yes, it is a cascade circuit, not a cascode. As I wasn't around in the days where many people had this sort of thing as a hobby (Im 24 now) and such, I learnt from older friends and older books. Ham radio/radio repair books from the 30s to the 50s were of big help to me. Things like the Radio designers handbook and such are a must have. Once I learnt what all the individual components did in a circuit I started to be able to piece things together and fault find in equipment. Of course i'm still learning all the time.

The consideration in the transformer to be used in the circuit needs to be of top quality. I will get my MEQ5 out and take some measurements. 
 
Thanks :) I'll looked at the cascade topology. Right now, I'm looking into a Mu-Follower device and guess it should be a nice candidate. As for the transformers, I always get some Cinemag when it's time to high quality transformers :)
 
So here is another tentavive with a Mu-Follower topology, made with a 6SN7 tube. Tell me if it's better, and my math ok!? Thanks guys :)
 

Attachments

  • mu-follower eqp.pdf
    887.1 KB · Views: 31
here is the MEQ amp, that 6AQ5 is a very nice sounding tube, largely overlooked (6BQ5 stole the show), so NOS versions are dirt cheap,

you could also use a 6V6GT (same tube as 6AQ5) and the octal socket is more rugged and easier to wire,

this amp sounds better than the EQP push pull and is easier to wire, does not need the 4 million dollar HS-29, and uses a wonderful sounding Triad OPT,

 

Attachments

  • MEQ-amp.jpg
    MEQ-amp.jpg
    212.1 KB · Views: 70
Deepdark said:
So here is another tentavive with a Mu-Follower topology, made with a 6SN7 tube. Tell me if it's better, and my math ok!? Thanks guys :)

It is better than the cascode since the 6SN7 has a lower plate resistance and can carry a higher current. The mu follower however is not a push pull topology so its current drive is half that of an SRPP with the same idle current. The 6SN7 has a low intrinsic distortion which the mu follower degrades very little. It will drive about 20V rms into a 10K load at about 0.4% distortion. To drive a 600 ohm load you need a 10k:600 transformer which will give about +16dBu into 600 ohms but the overall gain will only be 14dB - not enough. If you only plan to drive 10K bridging load then you could use a good quality 600:600 transformer and have the full 26dB gain available.

This is exactly the circuit used in my poor man's tube gain make up. It uses the 6CG7 which is the B9A version of the 6SN7.

PMTGMUcctjpeg


Cheers

Ian
 
Thanks Ian and CJ. Ian, i just saw that your poor mans and my design are really close. Can you explain a little about the impedance required to drive the unit at 26db. If it's possible to give me a little explanation. Thanks for helping me :)
 
Deepdark said:
Thanks Ian and CJ. Ian, i just saw that your poor mans and my design are really close. Can you explain a little about the impedance required to drive the unit at 26db. If it's possible to give me a little explanation. Thanks for helping me :)

Sorry, I don't understand the question - not sure what you mean by 'impedance to drive the unit at 26dB'.  Can you explain a bit more what you want?

Cheers

Ian
 

Latest posts

Back
Top