Isotope MM phono preamp by doors666

[jls001]

Some build pics of the Isotope MM phono preamp designed by our fellow forumer doors666.

Circuit details here: Isotope: Opamp based MM phono - diyAudio

Blank circuit board:
https://m.imgur.com/gallery/NgJ50xl

Board populated:
https://m.imgur.com/62CHDHT

Blues Mini Shunt Regulated Power Supply PCB:
https://m.imgur.com/IAnjTX6

Shuntreg board populated:
https://m.imgur.com/GGVqboX

https://m.imgur.com/D90Ha0c

Assembled in cabinet:
https://m.imgur.com/b9ZK4cd

Up and running:
https://m.imgur.com/2tipCLj

Slight deviations from the original design:

1/ the original design brief for the power supply is to use 18-0-18 secondaries. Since I have unused 22-0-22 R Core, I used that instead. No design changes necessary but a couple of resistor values need to be of higher wattage to take care of the higher heat dissipation.

2/ impedance loading resistor is tailored for my cartridge (~10x of internal resistance of the high output MC cartridge that I'm using). For a regular MM, 47k ohms would be ideal, and that is what comes with the kit.

3/ I have unusually high incoming AC supply voltage in my home (?238VAC at all times) so my secondaries are more like 24+ Volts instead of 22V. Though the designer had used 16V zeners to decide the final DC supply values, mine were more like +/-17.5V which is on the upper end of the supply voltage range for most dual opamps. Using this high value, my opamps over heated and went into thermal shutdown after playing about one and half LP sides. The solution was to use lower value zeners. I tried 12V and 15V and settled for 15V zeners (final DC voltages are +15.95 and -15.92V) as I like the sound more with 15V. There's no more thermal shutdown with this DC supply value.

4/ since I use unity gain buffer instead of a line level preamp with some gain, I felt that the gain of the Isotope was low, so with the help of the designer I have increased the gain a bit to be in line with other line level sources.

I have been doing some opamp rolling. Have tried NE5532, LME49990 and AD823. All of them sound very nice. There are definitely variations in sound but I haven't been able to decide yet what I like most. I'm yet to try my fav OPA2134 as one of the two I have is faulty.

I feel there are a couple of good reasons why this phono preamp sounds good (besides sound circuit design):

1/ the RIAA section of the preamp uses hand matched resistors and capacitors. The deviation from the RIAA curve is less than 0.2 dB from 20 Hz to 40 KHz.

2/ use of good quality parts.

I'll post listening impressions later. But for now, I'll say that it is a very, very good phono stage.

Heartfelt thanks to doors666 for coming up with such a good design, and for putting up with me and hand holding me in troubleshooting problems.

 

[jls001]

Some
I feel there are a couple of good reasons why this phono preamp sounds good (besides sound circuit design):

1/ the RIAA section of the preamp uses hand matched resistors and capacitors. The deviation from the RIAA curve is less than 0.2 dB from 20 Hz to 40 KHz.

2/ use of good quality parts.

I forgot the third and equally important reason - the use of a very good shunt regulated supply. Having used both the full sized and mini, I'm impressed with doors' shuntregs :thumbsup:

 

[jls001]

Reposting the photos:

Blank circuit board:

Board populated:

Blues Mini Shunt Regulated Power Supply PCB:

Shuntreg board populated:

Assembled in cabinet:

Up and running:

 

[hifipal4all]

Some build pics of the Isotope MM phono preamp designed by our fellow forumer doors666.

Circuit details here: Isotope: Opamp based MM phono - diyAudio

Blank circuit board:
https://m.imgur.com/gallery/NgJ50xl

Board populated:
https://m.imgur.com/62CHDHT

Blues Mini Shunt Regulated Power Supply PCB:
https://m.imgur.com/IAnjTX6

Shuntreg board populated:
https://m.imgur.com/GGVqboX

https://m.imgur.com/D90Ha0c

Assembled in cabinet:
https://m.imgur.com/b9ZK4cd

Up and running:
https://m.imgur.com/2tipCLj

Slight deviations from the original design:

1/ the original design brief for the power supply is to use 18-0-18 secondaries. Since I have unused 22-0-22 R Core, I used that instead. No design changes necessary but a couple of resistor values need to be of higher wattage to take care of the higher heat dissipation.

2/ impedance loading resistor is tailored for my cartridge (~10x of internal resistance of the high output MC cartridge that I'm using). For a regular MM, 47k ohms would be ideal, and that is what comes with the kit.

3/ I have unusually high incoming AC supply voltage in my home (?238VAC at all times) so my secondaries are more like 24+ Volts instead of 22V. Though the designer had used 16V zeners to decide the final DC supply values, mine were more like +/-17.5V which is on the upper end of the supply voltage range for most dual opamps. Using this high value, my opamps over heated and went into thermal shutdown after playing about one and half LP sides. The solution was to use lower value zeners. I tried 12V and 15V and settled for 15V zeners (final DC voltages are +15.95 and -15.92V) as I like the sound more with 15V. There's no more thermal shutdown with this DC supply value.

4/ since I use unity gain buffer instead of a line level preamp with some gain, I felt that the gain of the Isotope was low, so with the help of the designer I have increased the gain a bit to be in line with other line level sources.

I have been doing some opamp rolling. Have tried NE5532, LME49990 and AD823. All of them sound very nice. There are definitely variations in sound but I haven't been able to decide yet what I like most. I'm yet to try my fav OPA2134 as one of the two I have is faulty.

I feel there are a couple of good reasons why this phono preamp sounds good (besides sound circuit design):

1/ the RIAA section of the preamp uses hand matched resistors and capacitors. The deviation from the RIAA curve is less than 0.2 dB from 20 Hz to 40 KHz.

2/ use of good quality parts.

I'll post listening impressions later. But for now, I'll say that it is a very, very good phono stage.

Heartfelt thanks to doors666 for coming up with such a good design, and for putting up with me and hand holding me in troubleshooting problems.

Congrats Doors!!! :clapping:

Thanks jls001 for this useful post

Cheers

 

[doors666]

Thanx jls for starting the thread. I was feeling lazy to do so

Some build pics of the Isotope MM phono preamp designed by our fellow forumer doors666.

Circuit details here: Isotope: Opamp based MM phono - diyAudio

Blank circuit board:
https://m.imgur.com/gallery/NgJ50xl

Board populated:
https://m.imgur.com/62CHDHT

Blues Mini Shunt Regulated Power Supply PCB:
https://m.imgur.com/IAnjTX6

Shuntreg board populated:
https://m.imgur.com/GGVqboX

https://m.imgur.com/D90Ha0c

Assembled in cabinet:
https://m.imgur.com/b9ZK4cd

Up and running:
https://m.imgur.com/2tipCLj

Slight deviations from the original design:

1/ the original design brief for the power supply is to use 18-0-18 secondaries. Since I have unused 22-0-22 R Core, I used that instead. No design changes necessary but a couple of resistor values need to be of higher wattage to take care of the higher heat dissipation.

2/ impedance loading resistor is tailored for my cartridge (~10x of internal resistance of the high output MC cartridge that I'm using). For a regular MM, 47k ohms would be ideal, and that is what comes with the kit.

This is the most important aspect in an MM phono. If the loading is not correct, the HF response will go for a toss.

Those are hefty heatsinks, much smaller ones will suffice

3/ I have unusually high incoming AC supply voltage in my home (?238VAC at all times) so my secondaries are more like 24+ Volts instead of 22V. Though the designer had used 16V zeners to decide the final DC supply values, mine were more like +/-17.5V which is on the upper end of the supply voltage range for most dual opamps. Using this high value, my opamps over heated and went into thermal shutdown after playing about one and half LP sides. The solution was to use lower value zeners. I tried 12V and 15V and settled for 15V zeners (final DC voltages are +15.95 and -15.92V) as I like the sound more with 15V. There's no more thermal shutdown with this DC supply value.

The supply provides zener voltage + 0.7V as output voltage. I usually provide 10 zeners for zener matching. I was hoping to get around 16.5-16.75v out of it but the tolerance of the zeners made it much higher than that for you. I will probably provide 15v zeners in the kit henceforth as its a much more safe value.
Matching output voltage is not necessary for this application, but it should still be ballpark in same range. Running the rails at 14.5 and 15.5 isnt optimal, though it will still work.

4/ since I use unity gain buffer instead of a line level preamp with some gain, I felt that the gain of the Isotope was low, so with the help of the designer I have increased the gain a bit to be in line with other line level sources.

This can be easily changed by changing a resistor value. Standard gain doesnt take you all the way to 2V, so you might have to turn the volume knob a bit higher as compared to the usual 2v sources.

I have been doing some opamp rolling. Have tried NE5532, LME49990 and AD823. All of them sound very nice. There are definitely variations in sound but I haven't been able to decide yet what I like most. I'm yet to try my fav OPA2134 as one of the two I have is faulty.

I feel there are a couple of good reasons why this phono preamp sounds good (besides sound circuit design):

1/ the RIAA section of the preamp uses hand matched resistors and capacitors. The deviation from the RIAA curve is less than 0.2 dB from 20 Hz to 40 KHz.

2/ use of good quality parts.

I'll post listening impressions later. But for now, I'll say that it is a very, very good phono stage.

Heartfelt thanks to doors666 for coming up with such a good design, and for putting up with me and hand holding me in troubleshooting problems.

It does sound awesome, doesnt it. I also have my old cnc board, I installed a measured riaa into that, but no comparison. I also installed the same riaa (from lme49720 datasheet) with isotope board, while it was better than the cnc board, not as good as isotope riaa on isotope board.

Another thing you can try to do is try a few opamps and see which ones gives you minimal dc offset and then bypass the output caps and try. There will be some dc offset and your preamp should be able to handle that.

In the pics, I see that you have braided some of the wires, I have seen quite a few builds here using the same method. You are supposed to twist the wires, not braid them. The cancellation effects work with twisted pairs or triplets, not with braided.

 

[Hari Iyer]

Great DIY work.

I would suggest to add a jfet in the input stage before the op-amp with a constant current source. This will prevent loading of the cartridge and also the output swing will be very high due to the constant current source (CCS). The CCS will prevent any variation in output voltage due to the voltage swing. I am building this RIAA phono pre-ap currently with this particular topology which should be ready this weekend.

@ Jousha - it will be nice of you, if you can review my RIAA phono stage too if time permiits.

 

[Hari Iyer]

Attaching my rough sketch for your comments.

cheers,

 

[doors666]

Attaching my rough sketch for your comments.

cheers,

I dont see it. Could you please attach again.

 

[jls001]

Faceplate for the Isotope:

 

[prem]

Nice. I like the front plate

 

[Hari Iyer]

I dont see it. Could you please attach again.

Here you go. My phono pre-amp is based on this sketch with many components modified to suit availability and precision.

You can see the constant current source used in the jfet source which helps to retain the voltage changes across output of the jfets due to change in the input line voltage from the pickup. This superbly allows greater clarity than directly connecting the pick-up to the op-amp inputs. Also the 75micorsecond RIAA filer is not in the direct feedback loop of the RIAA which allows for a tighter bass response with this circuitry imo.

 

[jls001]

I've clocked many hours of play time as well as went through lots of experiments to optimise the sound.

First of all, I found the gain inadequate. So with the help of the designer I changed the two resistors responsible for gain from 120 Ohms to 100. This improved the loudness, but it still wasn't enough in my setup as my buffer is unity gain and my power amp is configured for fairly low gain (~27-28 dB). So I experimented further with 68 Ohms. This was way too loud, so I'm now using 82 Ohms which gives similar subjective loudness as other line level sources. Fortunately, I don't hear noise penalty after raising the gain.

I also rolled many opamps, namely NE5532 (stock supplied with kit), AD823, LME49990, and LM833.

I also had a nagging doubt the optimal resistive loading for my cartridge (high output Denon DL 160 MC cartridge) which is supposed to output about 1.6 mV into a 47K Ohm load. I've found long ago that 47K doesn't sound right for this cartridge on three different phono stages, so I've been using ~10x internal DC impedance of the cartridge (160 Ohms). I used the nearest standard value of 1.8K. Since there are quite a few users who like 1K loading, I tried that too but liked 1.8K much more.

So how does the Isotope sound? Even with a cold start it becomes obvious that the Isotope is very balanced across the whole frequency spectrum. There is no mid bass bump or spotlighted mids, or recessed/highlighted highs to spoil the experience. It is a very resolving phono preamp and very adept at digging out details buried deep in the mix, but its standout feature is its ability to portray a very holographic sound stage where instruments and voices have a specific place in a 3D space. Very impressive, I must say.

I've had it butt head with my other MM-capable phono preamp - the Pass Pearl 2 clone. This phono stage is a much more complex and elaborate design with mono boards populated with discrete parts only (no ICs), and an outboard power supply unit. The only thing swapped is the phono preamp.

The only area where the Pearl 2 betters the Isotope is in bass weight. In every other parameter the Isotope wins.

To come back briefly to the results of opamp rolling, I was quite surprised that I didn't hear much discernible change from one opamp to another. In fact optimizing the resistive loading of the cartridge had a much more telling effect on the sound. So I'm sticking to the NE5532 that came with the kit. I recall that when I had the CNC phono preamp, I liked the sound of the OPA2134 the most out of the opamps I tried on that preamp.

I also tried a Denon DL 103 LO MC via a step up transformer. I'll have to troubleshoot my SUT as there is a slight channel imbalance.

To conclude, the Isotope is a very fine MM phono preamp. I like it a lot!

 

[Hari Iyer]

The isotope is directly coupled to the opamp inputs and this could probably be the reason for low gain. I would recommend something like a jfet before the opamp which can offer excellent coupling to the isotope. Combined with a CCS it would be a winner imo.

 

[omishra]

The isotope is directly coupled to the opamp inputs and this could probably be the reason for low gain. I would recommend something like a jfet before the opamp which can offer excellent coupling to the isotope. Combined with a CCS it would be a winner imo.

Hari, some of the opamps he used are jfet input stages. Need no extra jfet as they already tend to exhibit large input impedance.

 

[jls001]

@Hari: doors666 will be abler to reply to your suggestion as he is the designer. However, from my observations and experiments, the two gain stages can be tweaked as desired by simply playing around with resistor values on the feedback networks. In fact as I had noted in my post above, it was possible to get too much gain.

In my limited experience, more than the circuit topology, it is correct resistive and capacitive loading AND rigorous cartridge setup that bears greater fruit (assuming generally decent electronics chain). Last night I used a mirror to check the azimuth alignment of my cartridge and found that it was slightly off. After correcting it, separation improved. I plan to re-check the mounting distance, offset angle, over hang, tracking angle and even the tracking weight. It's a tedious job but the results can sometimes be stunning. Fine tuning mechanical parameters in TT playback is critical.

 

[Hari Iyer]

Hari, some of the opamps he used are jfet input stages. Need no extra jfet as they already tend to exhibit large input impedance.

Jfet input op-amps are not good at low level resolution imo. i would rather prefer bjt input ones. Discrete jfets are better if they use a CCS as per my above schematic n post #11. Correct me if i am wrong.

 

[doors666]

I've clocked many hours of play time as well as went through lots of experiments to optimise the sound.

First of all, I found the gain inadequate. So with the help of the designer I changed the two resistors responsible for gain from 120 Ohms to 100. This improved the loudness, but it still wasn't enough in my setup as my buffer is unity gain and my power amp is configured for fairly low gain (~27-28 dB). So I experimented further with 68 Ohms. This was way too loud, so I'm now using 82 Ohms which gives similar subjective loudness as other line level sources. Fortunately, I don't hear noise penalty after raising the gain.

its possible. The denon is a mc cartridge, and it has quite low output (though higher than a lot of other mc carts). The default 120 or 100R works fine with normal mm phonos, and the gain can be increased further by reducing those resistors. In my setup, I run it with 120R resistors and my volume control is roughly aroud halfway for decent listening.

I also rolled many opamps, namely NE5532 (stock supplied with kit), AD823, LME49990, and LM833.

I also had a nagging doubt the optimal resistive loading for my cartridge (high output Denon DL 160 MC cartridge) which is supposed to output about 1.6 mV into a 47K Ohm load. I've found long ago that 47K doesn't sound right for this cartridge on three different phono stages, so I've been using ~10x internal DC impedance of the cartridge (160 Ohms). I used the nearest standard value of 1.8K. Since there are quite a few users who like 1K loading, I tried that too but liked 1.8K much more.

This is very important, if the loading isnt right, the sound wont be good. Its even more important for mm carts.

So how does the Isotope sound? Even with a cold start it becomes obvious that the Isotope is very balanced across the whole frequency spectrum. There is no mid bass bump or spotlighted mids, or recessed/highlighted highs to spoil the experience. It is a very resolving phono preamp and very adept at digging out details buried deep in the mix, but its standout feature is its ability to portray a very holographic sound stage where instruments and voices have a specific place in a 3D space. Very impressive, I must say.

I've had it butt head with my other MM-capable phono preamp - the Pass Pearl 2 clone. This phono stage is a much more complex and elaborate design with mono boards populated with discrete parts only (no ICs), and an outboard power supply unit. The only thing swapped is the phono preamp.

The only area where the Pearl 2 betters the Isotope is in bass weight. In every other parameter the Isotope wins.

That sounds like music to my ears. Pearl 2 from nelson pass is considered pretty much the gold standard for the phono preamps. I knew isotope is a very good phono preamp, but I never thought it will turn out to be better than pearl 2. I unfortunately have never heard the pearl.

To come back briefly to the results of opamp rolling, I was quite surprised that I didn't hear much discernible change from one opamp to another. In fact optimizing the resistive loading of the cartridge had a much more telling effect on the sound. So I'm sticking to the NE5532 that came with the kit. I recall that when I had the CNC phono preamp, I liked the sound of the OPA2134 the most out of the opamps I tried on that preamp.

I also tried a Denon DL 103 LO MC via a step up transformer. I'll have to troubleshoot my SUT as there is a slight channel imbalance.

To conclude, the Isotope is a very fine MM phono preamp. I like it a lot!

Even i didnt hear too much of differences with different opamps. Maybe very slight changes, and it will at the end boil down to personal preference.
thanx for the reviews and comparisons...

The isotope is directly coupled to the opamp inputs and this could probably be the reason for low gain. I would recommend something like a jfet before the opamp which can offer excellent coupling to the isotope. Combined with a CCS it would be a winner imo.

Gain is controlled by feedback resistors. Isotope uses passive riaa network, so its easy to change the gain. Your design uses active riaa network, so its harder there. You can change the gain in your first stage as its feedback network doesnt have riaa components. The jfets and the ccs are part your input stage and that wont have any effect on the gain.

I have seen your schematic. Its from Bob Cordells LSK489 appnotes. Most of the jfet input opamps have the same input stage as yours, with the ccs. The bjt input opamps also have similar input stages, differential with ccs. I would have preferred if you had gotten rid of the first opamp and used a discrete gain stage also. Which jfets are you planning to use for this. If i remember correctly, lsk489s were quite expensive last i checked. Have you tried to simulate the circuit, how's the response, how much is the gain?

 

[Hari Iyer]

Gain is controlled by feedback resistors. Isotope uses passive riaa network, so its easy to change the gain. Your design uses active riaa network, so its harder there. You can change the gain in your first stage as its feedback network doesnt have riaa components. The jfets and the ccs are part your input stage and that wont have any effect on the gain.

Having less drop in the input stage is critical imo for the later stages. In the ckt posted by me, you can control the gain by increasing or decreasing the ccs. In this case the 1K resistor connected to the emitter is the key to change the gain of the output stage & not only the feedback resistor from the opamp to the gate. This resistor controls the input gain which is independent of the op-amp gain. Infact the ccs is used more to allow the feeding of the input signal to the op-amp without degrading the signal due to output voltage swing due to low frequency gain that will be required for the op-amp.

I have seen your schematic. Its from Bob Cordells LSK489 appnotes. Most of the jfet input opamps have the same input stage as yours, with the ccs. The bjt input opamps also have similar input stages, differential with ccs. I would have preferred if you had gotten rid of the first opamp and used a discrete gain stage also. Which jfets are you planning to use for this. If i remember correctly, lsk489s were quite expensive last i checked. Have you tried to simulate the circuit, how's the response, how much is the gain?

I have used equivalent jfets instead of LSK489 with excellent results - mine is 2N2145 smds which is close to the LSK489s. I have used a tank circuit for input cartridge loading to reduce input cable & RF/EMI noise pick-up. The first op-amp is used as a feedback loop for 75microsecond and the second opamp is used for 318 & 3180 microseconds. There are some errors in the component values which needs to be corrected in the above schematic.

You need to add many other power conditioning filters to the above schematic. Also input and output impedance, noise filters needs to be added for the above ckt to work perfectly. Unless you are knowledgable in that area, the ckt will just not work. I have used a DCB1 at the output stage of this RIAA with excellent results using equivalent smds.

 

[doors666]

yeah you are right about the gain of the input stage, i didnt notice that you are taking the output from the drain, its wired as a gain stage not as a buffer. By varying the ccs current, you can control the gain.

can you post the actual schematic. its difficult to talk about a circuit when the full schematic is not there.

whats the part number for the jfet. i cant find a jfet called 2n2145. much less its smd dual. There are not many jfets that can equal the lsk series (these are basically mostly copies of toshiba 2sk170 and its duals once those were out of production). The only jfet that comes close in performance to 2sk170 (and its variants) is BF862 and i dont think it has a dual.

you have a tank circuit as in an LC network in your input stage, with an actual physical inductor? Or are you talking about the LC formed by the cartridge inductance and loading capacitance.

 

[Hari Iyer]

yeah you are right about the gain of the input stage, i didnt notice that you are taking the output from the drain, its wired as a gain stage not as a buffer. By varying the ccs current, you can control the gain.

can you post the actual schematic. its difficult to talk about a circuit when the full schematic is not there.

whats the part number for the jfet. i cant find a jfet called 2n2145. much less its smd dual. There are not many jfets that can equal the lsk series (these are basically mostly copies of toshiba 2sk170 and its duals once those were out of production). The only jfet that comes close in performance to 2sk170 (and its variants) is BF862 and i dont think it has a dual.

you have a tank circuit as in an LC network in your input stage, with an actual physical inductor? Or are you talking about the LC formed by the cartridge inductance and loading capacitance.

The noise filter is not actually a tank ckt so to speak, but two 220 ohm resistors in series. The center point of this resistors is connected to a 470pf capacitor which is grounded. It works like this - the input signal + noise sees a resistance of 220 ohm and then another 220 ohm. The signal passes through the series resistance. But the noise being random and weak finds the second 220 ohm resistance too high and instead finds an easier path to ground through the 470pf capacitance. This is my own invention. The noise just vanished with this. I am thinking of increasing the second 220 ohm to 470 ohm.

You are not correct about the LSK485 and 2SK170BL. The 2SK170BL has higher input capacitance which resonantes at around 18KHz and is not suitable for any phono preamp. The LSK485 has a ultra-low input capacitance which prevents the resonance of the cartridge at 18KHz and hence is suitable for the phono pre-amp inputs. The removal of this resonance gives an excellent response to the pre-amp. This is explained in the datasheet as well. As rightly pointed out the LSK485 is very very expensive and is rare to find.

My bad the replacement for LSK485 is 2SK2145. The datasheet link is here 2SK2145 | MOSFETs | TOSHIBA Storage & Electronic Devices Solutions Company | Asia-Pacific
I have found this jfet 2SK2145 as an excellent replacement for the LSK485 and discovered this after a lot of R&D.

My schematic is hand sketched and needs to be formatted for proper view. Will do so in the weekend and then post.