I've been using a strain gauge cartridge for a couple of years now and have been really happy, but I have an urge to go back and listen to my ZYX cartridge. Only thing is, my phono amp is a valve unit (Allen Wright VSE pre) that doesn't really have enough gain. So I'm looking at building a new phono amp before I swap cartridges.
Target topography includes:
1. All balanced
2. 1 SUT and 2 amplification stages
3. Passive equalisation.
4. High gain FETS where possible
I have a Jensen SUT (10x) but I still want max voltage gain in the first stage. I've been looking at a few ideas and like the microphone pre-amp I saw on the Rod Elliot site. See Figure 9 in the attached article.
Compound vs Darlington
I've built this compound circuit with a 2SK369 and BC559 as per circuit diagram attached. I couldn't sim this exact circuit in Simetrix because it doesn't have a model for the 2SK369 but I did a sim using a J110 fet and simulation shows that I can expect a gain of 150x. Much better than I would get using the Fet by itself in a normal grounded source configuration.
So I just fired this up and - wow, with an input of 5mV, I measure an output of 8.5V!!! So a gain of about 1,700x. Much than I expected and much more than I need.
What do you guys think? Is this sub-circuit worth further investigation? I can lower gain by
1. putting in source resistors - say 10ohm for each fet
2. easier way is putting a resistor across the drains. But this will interfere with the eq network between stages. But it would be a cracking ide if this was a flat gain stage.
Any suggestions will be considered as long as they don't include valves! I love valves, I already have a valve pre, but this is gonna be a s-s job.
Update 4 April 2016
Member Polyphaze has solved my questions about the first stage. Further discussion on the complete circuit is included in new thread. Please feel free to participate.
http://www.diyaudio.com/forums/anal...obos-balanced-mc-phono-stage.html#post4670610
Target topography includes:
1. All balanced
2. 1 SUT and 2 amplification stages
3. Passive equalisation.
4. High gain FETS where possible
I have a Jensen SUT (10x) but I still want max voltage gain in the first stage. I've been looking at a few ideas and like the microphone pre-amp I saw on the Rod Elliot site. See Figure 9 in the attached article.
Compound vs Darlington
I've built this compound circuit with a 2SK369 and BC559 as per circuit diagram attached. I couldn't sim this exact circuit in Simetrix because it doesn't have a model for the 2SK369 but I did a sim using a J110 fet and simulation shows that I can expect a gain of 150x. Much better than I would get using the Fet by itself in a normal grounded source configuration.
So I just fired this up and - wow, with an input of 5mV, I measure an output of 8.5V!!! So a gain of about 1,700x. Much than I expected and much more than I need.
What do you guys think? Is this sub-circuit worth further investigation? I can lower gain by
1. putting in source resistors - say 10ohm for each fet
2. easier way is putting a resistor across the drains. But this will interfere with the eq network between stages. But it would be a cracking ide if this was a flat gain stage.
Any suggestions will be considered as long as they don't include valves! I love valves, I already have a valve pre, but this is gonna be a s-s job.
Update 4 April 2016
Member Polyphaze has solved my questions about the first stage. Further discussion on the complete circuit is included in new thread. Please feel free to participate.
http://www.diyaudio.com/forums/anal...obos-balanced-mc-phono-stage.html#post4670610
Attachments
Last edited:
Why not simply build a head amp (pre pre amp) to increase the ZYX's signal to MM levels, so your VSE phono stage will be able to handle it?
I have one of Hugh Dean's 'Paris' head amps (but I'm not sure whether he still offers it) and there is the BoozHound Labs pre pre kit.
Andy
The current sink is a cascode CCS as per Gary Pimm. Can't find Gary Pimm web site right now or else I would attach link here.
Thanks Andy, I could go this route, or even use my Jensen SUT in front of the VSE (I love the sound of the VSE and I think I would prefer a transformer to an active solution). But, the rest of my system is balanced, including my strain gauge pre-amp. I want to build something that is mine, and I want it balanced.
Ha ha ha, there is nothing new under the sun, is there! Your circuit is, conceptually, very very similar to what I am building, except that you have the compound circuit as second stage and I am using it for the first stage. But maybe your design is the better way to go. I am using 2SJ74 p-jfets for the second stage, and I am worried that I will overload the input if I am swinging too much voltage out of first stage. Here is my complete circuit diagram I've built and tested 1st gain stage and eq network - I want to finish 2nd gain stage over the weekend). How much gain do you get out of each stage? What cartridge do you use, and what is its output? It looks like you get enough gain without a SUT. Do you get a full 2V output?
Attachments
There is more than enough gain available in this circuit. As shown, open loop gain is 230x in the first stage, and 75x in the second, for a total of 17250x. With EQ added, at 1kHz, gain is 1800x, or 65db. Halving the emitter resistors in the second stage would boost it to 70db. 65db is just right for my .3mV Denon.
IMO a bjt is a better input device for an MC. However, I do like the idea of a fet in the second stage compound. I'm going to try that.
Many years ago I built a tube phono stage based on the front end of Allen Wright's RTP. With Allen's blessing, I used a MAT02 bjt at the input cascode instead of the 2SK147 fets. The result was alot more gain and whisper quiet. It is still my favorite phono stage, and the one I use the most.
IMO a bjt is a better input device for an MC. However, I do like the idea of a fet in the second stage compound. I'm going to try that.
Many years ago I built a tube phono stage based on the front end of Allen Wright's RTP. With Allen's blessing, I used a MAT02 bjt at the input cascode instead of the 2SK147 fets. The result was alot more gain and whisper quiet. It is still my favorite phono stage, and the one I use the most.
Thats great, thanks Polyphase. Now I can see where the family resemblence comes from - we are both using Allen Wright RTP as a starting point. I have his single end VSE FVP pre-amp, but never had enough cash for the RTP. So my circuit above is my cheap solution using s-s devices. But, the more I look at your Phobos, the more I like it.
One advantage we have with s-s is the p-channel device. We can make compound devices that provide huge amounts of gain stage. But we can also do away with inter-stage coupling caps. If we cascade an n-channel stage into a p-channel stage, like my circuit in my last post, then there is no need for coupling cap. So if I took the Phobos as a staring point and modified it slightly, then it might look like this.
What do you think?
Attachments
Poly, a PCB would be great. I will send PM with details. I've ordered a pair of MAT02, I was lucky to find 2 devices local in Australia (well I will be lucky if they are not fake). I will also have a go at working out some circuit values.
Do you have a design thread for the Phobos already? I did a search and couldn't find one. This has rapidly moved from a discussion of the first stage to the complete pre-amp so a new thread makes sense.
Regards, Hazard
Hi,
the PNP-type MAT03 is slightly lower in noise than the NPN MAT02.
It ´d also allow for the useage of N-channel JFETs in the second stage which are easier to souce than P-channels.
As alternative for the MATs there´d be the 300-series from THATcorp.
THAT also offers matched Quad which allowed to parallel the input transistors to reduce noise and increase gain.
Since the MC-stage is made basically from two singleended-amplifiers the noise also increases by 3dB against a singleended stage.
Since the output of #7 is a rather high impedance current output a Buffer stage might be needed.
The Phobos ´s has a low impedance output, therefore a buffer is obsolete here.
jauu
Calvin
the PNP-type MAT03 is slightly lower in noise than the NPN MAT02.
It ´d also allow for the useage of N-channel JFETs in the second stage which are easier to souce than P-channels.
As alternative for the MATs there´d be the 300-series from THATcorp.
THAT also offers matched Quad which allowed to parallel the input transistors to reduce noise and increase gain.
Since the MC-stage is made basically from two singleended-amplifiers the noise also increases by 3dB against a singleended stage.
Since the output of #7 is a rather high impedance current output a Buffer stage might be needed.
The Phobos ´s has a low impedance output, therefore a buffer is obsolete here.
jauu
Calvin
Thanks Calvin, I didn't know about MAT03, but I already have a few 2SJ74 so I am good for p-channel fet. I was able to buy MAT02 on eBay so I have all parts I need.
I was just about to respond to your comments about Phobos but I see that you have already picked this up in your next post. In practice, I am very comfortable with taking the output from the drain (high output node) of the output. My strain gauge pre-amp is a very simple circuit - a differential amp made of two 2SK369 set up as LTP. I take the output across the drains of the fets. I don't use a pre-amp in the normal sense, I adjust gain by switching resistors across the drains - this reduces the gain of the pre-amp stage. The output goes straight to the input of my power amps which have j-fet input and 1Meg input resistance. The output of the Phobos can be set up in exactly same way.
Here is the circuit for my strain gauge pre-amp. Output can go via capacitor, I have done it this way, but currently I am using transformer coupling (Lundahl LL1527) between the pre-amp and power amp. The primaries span the drains, and the secondaries go to power amp inputs (power amp is fully differential).
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.