Tweaking a Phono Preamp

[Frankentronics]

Greetings,

It took a while, but I finally assembled a phono preamp based on the following BoozHoud Labs design.



I made the circuit home brew style, like this.



I first tested with an 18V wall wart (which produced hum) then an 18V battery and wasn't getting enough volume. I thought a higher voltage would produce more gain, so I used a salvaged 24V power supply (it is raised 3/4" from the metal enclosure). There is no hum, but the volume is still too low. Below is an image of the project in its (salvaged) enclosure.



Since the gain is too low I am now thinking that the 2SK170 JFETs are not what they are supposed to be (I got them through Amazon). In another thread many folks told me where I can get reliable 2SK170 or LSK170 JFETs, so I do not need help with that. I just don't want to buy new JFETs if the existing ones are not the most likely source of the problem.

Any advice would be appreciated.

Thanks...

[chemelec]

You Might Build this Preamp:
http://chemelec.com/Projects/Preamp-1/Preamp.htm

[Benta]

The transconductance spread of the 2SK170 is enormous (like 1:10). The 50 ohm source resistors help, but you'll still have a large spread.
The original Toshiba parts could be ordered as selected parts in three transconductance ranges. If you can find the highest grade somewhere, it might solve the issue (2SK170V).
Messing with the supply voltage will not help.

[Conrad Hoffman]

It's not uncommon for RIAA stages to need some sort of follower amp to get enough gain. Is the gain of the two channels the same? That would suggest that the FETs are at least the same.

Entirely IMHO and FWIW, I think the performance of RIAA stages with just a couple transistors or FETs is inherently limited by available gain. My preamp of choice is based on the old app note from when the LME series opamps came  out. Passive eq is also a solution for a problem that probably doesn't exist and one shouldn't be too afraid of including it in the feedback.

[b_force]

Capacitors parallel to R3 and R12 might help.
Since the values of the resistors are pretty low that's gonna be in the uF range.

You can calculate it with f=1/(2 pi R C)

[cvanc]

Capacitors parallel to R3 and R12 might help.
Since the values of the resistors are pretty low that's gonna be in the uF range.

You can calculate it with f=1/(2 pi R C)

Agreed; it's an easy "retrofit" to your board that should bring up the gain.  Try it as an experiment and let us know what happens.

(and I hope you're having fun with this, your board looks nice)

[Frankentronics]

Capacitors parallel to R3 and R12 might help.
Since the values of the resistors are pretty low that's gonna be in the uF range.

You can calculate it with f=1/(2 pi R C)

If I understand correctly, I really need to calculate the capacitance, like so: C=1/(2 pi R f), correct? But what value do I use for the frequency f?

[Frankentronics]

I did some Googling and reading and I think I have a better understanding.

As I understand it, the frequency in the equation is the desired cutoff frequency, which is usually around 20Hz. I found a practical Online Calculator.

If I want the cutoff frequency to be 20Hz I would need 160uF caps there. I happen to have some 150uF caps, so the cutoff frequency will be 21Hz.

So, assuming that's all correct I will do some testing tomorrow and post the results.

Thanks all for your help.

[b_force]

I did some Googling and reading and I think I have a better understanding.

As I understand it, the frequency in the equation is the desired cutoff frequency, which is usually around 20Hz. I found a practical Online Calculator.

If I want the cutoff frequency to be 20Hz I would need 160uF caps there. I happen to have some 150uF caps, so the cutoff frequency will be 21Hz.

So, assuming that's all correct I will do some testing tomorrow and post the results.

Thanks all for your help.

Compliments you did some research youself
Yeah around 10-20Hz is fine. Since 220uF is a standard value, I would take that.
I wonder why such low value resistors are being used.
Experiment a little you might only need one of them.

For people who are looking for a similar circuit, I would recommend using opamps instead.
Easier, smaller and better performance.
Otherwise some old tubes are classy of course.

[sd]

What kind of cartridge do you have, MM, MC, MC high gain?

[Frankentronics]

Hello again, and thank you for your replies.

To answer your questions...

I'm using a Bang & Olufsen BEOGRAM RX2 turntable that I found in the garbage (didn't need any work). It came equipped with MMC3 cartridge, which is an upgrade.



It is hooked up to the AUX channel of a Sherwood RX-4103 that I also found in the garbage and feeding into a pair of 8ohm speaker that I found in the garbage.

At some point I might switch to a Technics SL-B20 that I found in the garbage (needed a new belt) and also has an MMC.



You can also see on my photo a TEAC A-4010S that I bought for $50 on Craig's List and that needed a new belt.

I love NYC.


I performed a test with 220uF caps as well as 150uF caps. I can't say that there's a sufficient difference. There is still a noticeable difference in volume between the turntable and any other source. Also, the sound is simply not very rich.

I might build another preamp, either with op amps or transistors (perhaps the one chemelec suggested, here), but I still want to spend a little more time troubleshooting this one. I also like the idea of JFETs, as they are supposed to perform more like tubes. Which brings me to a question, would there be an opamp that anyone recommends that has internal JFETs?

So, what's my next experiment with this one? Buying two pairs of LSK170X-1 or 2SK170BL from a reputable source?

Thanks...

[cvanc]

Hmmm.

Do you have a scope, can you look at your signals?  How about a tone generator to test with continuous sine waves?

Also, you might want to test using your other turntable?  It would help to know if the problem stayed the same.   I suppose this might be a problem with your current source (cartridge or stylus issue, bad tonearm wiring, etc)  Or maybe use a different line input on your amp?

[Alex Nikitin]

If you are OK working with static-sensitive devices, you may try to assemble my phono stage. It is also FET only, passive EQ circuit  . To my ear it is better than the Pacific and you can get the FETs from a reputable source, like Mouser or Digikey. You can replace M4 with a 10 Ohm resistor for simplicity.

Cheers

Alex

P.S.  - if you are not sure about working with small MOSFETs, here is the circuit with BJTs and a JFET.

[Frankentronics]

Thank you both for your replies.

I do have a 20MHz scope that works OK (needs alignment, verticals are skewed), but I do not have a function or audio generator. I've been wanting to put together one but my previous attempts were not very successful. I'm still looking for a good schematic, I guess.

I tried both turntables and results are the same.

There is a boost in volume, but there's still a difference when I switch to CD or tape. I already tried a different input channel and results are the same. So, it is definitely the preamp itself that needs tweaking.

Perhaps those JFETs are really not as advertised.


Now I am also intrigued about Alex's phono preamp. I read the whole thing and wouldn't mind giving that build a try.

Do you have a schematic of the power supply, BTW?


Thanks...

[Kleinstein]

The amplifiers are already in a range where r3 is rather small and the trans-conductance is responsible for limiting the gain. To it does not really help much to parallel R3 with a cap - this adds a little gain, but also to the THD.

The circuit is simple, but also limited performance and dependent on the FET parameters that can scatter quite a bit.

[Frankentronics]

I was in fact hearing distortion when I used the modified circuit, later during the day.

So, this is basically not a very good preamp, I guess. I also noticed that Boozhoud Labs now modified this circuit by adding an additional gain stage, and an output buffer. Perhaps they had a few complaints and they had to make it louder.

Now I'm thinking I should probably just build another circuit and I'm quite intrigued by the circuit Alex posted.

[Conrad Hoffman]

As I alluded to earlier, phono stages with just a couple devices have to decide what to do with the limited gain- use it for , well, gain, or use it in the form of feedback to get decent performance. Tough to have both, thus the need for output stages with gain. If you look at many vintage receivers, they were able to use simple RIAA stages for that reason. IMO, if you like to mess with phono stages, you should build an inverse RIAA network to test them. I use this one- http://www.hagtech.com/pdf/riaa.pdf

[b_force]

Yes, the type of cartridge is also important, since the output voltage can be significantly lower.
Also there is some input RC network missing at the input of this circuit.

If you want a good, simple and cheap performing RIAA pre-amp, just build it around opamps.
You can get away with a single opamp per side, although a double stage gives you a little bit more freedom.

If you want it nice and fancy, go for tubes.

Everything in between is a bit of useless in my opinion.

[Alex Nikitin]

Now I am also intrigued about Alex's phono preamp. I read the whole thing and wouldn't mind giving that build a try.

Do you have a schematic of the power supply, BTW?


Thanks...

Any reasonable power supply from 24 to 30V DC will be fine, a battery supply would be the best, however even a simple non-regulated supply will work OK.

Cheers

Alex

[Conrad Hoffman]

Here's the RIAA circuit I use, based on the old LME application note. The sim is with 1028 opamps, but those area  bad choice. The actual preamp uses LME49710 devices.

[Conrad Hoffman]

Long ago there was a NYC company, Ace Audio. They sold a preamp kit with this simple RIAA circuit. It worked well and I cloned it using many other transistors. This, and the above, are strictly for MM cartridges. For MC you need way more gain.

[TimFox]

The 2-jfet circuit in the original post is similar to two-triode tube preamps.
If you want accurate RIAA response, the passive network used here is better than feedback around the two devices due to low open-loop gain and the wide range of gain vs. frequency for RIAA.
You must include the output impedance of the first stage in the network, but with the high output impedance of the naked jfet, that’s dominated by the drain load resistor.  Vacuum triodes have lower output resistance.
Accurate equalization with feedback works well with an op amp, which has very high open-loop gain at low frequency, where the RIAA stage needs it.

[Frankentronics]

Thank you all for your replies.

Alex, I think I misunderstood one detail on your schematic because I saw -24V. I thought that meant there would need to be a power supply that delivers +24V and -24V.

Also, I don't understand what are R_CH_1, R_CH2 and R_CH_3.


Conrad, I studied your schematic but can't be sure if I understand some details. It's about the V1, V2 and V3. Is V1 the signal input? And are V2 and V3 power inputs of 15V?



Since I still like the idea of using JFETS I'm still looking at the newer version of the BoozHound preamp. Obviously they've added 3 more JFETS and they possibly did that because there was not enough gain.

So, looking at their revised circuit, is it worth building? I mean, does it look like the 2 added stages might actually solve my problem of not enough gain? If so, this might be a solution in my case, since I've already built their old circuit, and I'd just be adding to my existing build.

[Audioguru]

I replaced the Jfets phono preamp in my 1964 HH Scott stereo receiver with opamps that produce much less distortion. I haven't played my "record player" for about 30 years. Why are you??

[Alex Nikitin]

I haven't played my "record player" for about 30 years.

  

I replaced the Jfets phono preamp in my 1964 HH Scott stereo receiver with opamps that produce much less distortion.

That is probably why  .

Cheers

Alex