My phono preamp for MC cartridges, His Master’s Noise, was published as a feature article on diyAudio.com. Check it out there! I discuss overall system design, noise calculations, design for low noise, and present a circuit with excellent linearity, high overload margins, good conformance to RIAA equalization, high power supply immunity, and high stability.
Soon to come, a moving magnet preamp using the same balanced-in-unbalanced-out approach as His Master’s Noise.
I would love to get a better idea of your “soon to come moving magnet balanced to unbalanced phono preamp”. What a beautiful acronym: STCMMBUTPP.
The difference in noise level for an mm preamp with shorting plugs vs. anything else plugged in is always an interesting sight. So yes, please give me a balanced connection from the cartridge to the preamp input. I use both MC and MM cartridges and your HMN won’t do both obviously.
Hans, I’ve built and discarded three versions so far- they all worked, but two required parts that most people can’t get (e.g., 2SJ74) and one required some tricky adjustment. I *think* I have the right answer now… but you can count on balanced inputs.
Whiling away the days between the HMN and your MM design I have played with my own creation using a CineMag CMMI-2C and a triode strapped 6J9P. It does a few things right, but I don’t have the experience to avoid all pitfalls involved in radical departures from the norm. There must be better parts choices, if not lower costs, just around the corner.
Stuart, thank you for your response. I wouldn’t mind building with a couple 2SJ74 transistors. I have a few that are known to be real rather than fake. Using them in a balanced MM phono stage seems like a good idea compared to a tricky adjustment.
I’m just about to fire up my build of your HMN MC phono stage using the boards that jackinnj made available.
Best of luck! I saw a lovely build recently from a fellow who did one for himself and one for his company. Made me rather envious!
Do you have any pictures of those two HMN builds? I assume they were done using the pcb’s.
I am still struggling with getting heat sinks properly mounted to the parts that need them. There isn’t much room for them in a few places. Any visuals of successful attempts would be much appreciated.
Hans, everything is packed away at the moment (we’re in the process of moving), but I’ll try to post some pictures after we’re arrived and unpacked. One clue: only one of the transistors in each CCS might need a heatsink, the one with most of the voltage across it. And the heatsinks needn’t be very big.
I have one question about the His Master’s Noise phono Preamp power supply.How to calculate the the high voltage output,base on normal serial regulator,the Vout=Vref*(1+R1/R2),The R1 is +260v output 220K and R2 is 8K2,The output should be 192V,not +260V,So please help me to understand it.You comments will be great appreciate it.
Xiaojun- Because the hfe of the error amp transistor can vary as can the reference voltage, the regulator may need trimming to hit the desired output (these were the values I ended up with in my build). You can trim the output voltage on test for your particular build by paralleling another resistor with the 8k2 if your output is too low, or across the 220k if your output is too high.
Dear SY,thank you so much for you input.
About one year ago i built your “His Masters Noise RIAA Preamp and i am happy with it, thank you! Now since i am pensioner i would like to build something complete different (also RIAA). I have some Parts for long, but they suggest to use a Western 437A for the outputstage. This Tubes are so crazy expensive, i can and will not buy this. But i have WE 436A (Tetrode). I know you are specialist in connecting Tubes to Triode. Do you think i can use this connected to Triode with a good Result?
Thank you so much for your Help,
Kind Regards Walter
Guten Tag, Walter! Just for clarity, are you proposing to use the 436 in the second gain stage or for the cathode follower?
Hi Dear Stuart, I got a board from a Diy member I can not remember, I forgot completely and right now I will finish the project, but I noted that in the board is place for IR830 that is a different animal to the original one that come with a deplection mode Mosfet.
Could you send me some information if is available in your knowledge about this boards ?
I am Diyaudio Member my nick is dady.
I living in Spain
Ola Esteban! I would guess the board was from jackinnj from diyAudio. I think the CCS MOSFETs were still depletion mode but the IRF830 might have been used for the regulator pass element. Jack still participates at diyAudio and might have more information.
The two updates I would make: reduce the current for the first stage (D3a) to 10mA and the second stage (ECC88) to 5mA. Not critical, but the unit will run cooler and performance will be just as good.
Stuart, I loved your article and I built your phono preamp, and am starting to trim the RIAA. You say to set the buildout resistor R7 so that the 50Hz response is 20dB above the 1kHz response. Why is that? I thought the 50Hz response should be about 16.95dB above the 1kHz response. Even the 20Hz response is only 19.3 dB above.
Also, I assume that you meant to adjust C2 for the 20kHz response. C4 is the 2.2uF output cap.
Thanks for a great design.
That’s a typo- the 1kHz gain should be about 20dB lower than the 20Hz gain (19.3dB to be more exact, as you said). Thanks for catching that!
Appreciate the kind words- I still use this preamp all these years later and have had little urge to change for MC.
Thanks for the clarification. That definitely makes things easier. Can’t wait to start using it.
Let me know how you like it once it’s up and running!
I am pulling my hair out. I can’t get my hmnpp is work. I am bringing up the voltage with a variac. Everything looks good and then slowly dc appears at the output. MVolts at first and then it goes runaway, full out b+ 305 volts. Dangerous b+ on the gnd – I’ve been zapped. What is it a bad? MOSFET? I’ve changed them once.
You likely have a wiring error- the output is capacitively coupled, so there should be essentially zero offset.
I consider myself a beginner and have now for more than 40 years. I finally got my/your hmnpa up and running.
It’s really wonderful and i learned a lot. I bought the sowter 1480 which is a newer version of the 8055 i think. My cartridge is a spu at 2 ohms and the secondary resistance of the 1480 is 100 ohms . What would be the best value of the 6.8k resistor in this case?
One note i want to tell you about. Early on in the project I bought a set of pcbs from a fellow diyaudio guy, frank wilker. He had sold them as working and needing a case etc. Well his heater supply was terribly wrong, he had installed the diodes on the negative supply backwards. So I didn’t pick this up and had only tested it on the positive side. And so when I had it all ready to go the things worked initially but with in a minute the whole chassis went to 270+ volts, and threw a lightning bolt,- man that was the scariest and stupidest thing I have done in all of my lectronics life. I took some time off and beside still wanting to knock someones teeth out I am back building again.
I send you some pics it looks good too!
The loading resistor on the secondary is something that you’ll have to determine experimentally. If it sounds good as-is, then don’t sweat it too much.
Congrats on the build! It’s not an easy project and most people don’t get across the finish line.
C12, C16 – 0.1uF, whats best type of capacitor to use here, ceramic, MKP etc?.
Cant find much info on this outside of datasheet.
I’d use metallized polypropylene, cheap, easy to get, excellent mechanical and thermal characteristics, high performance. Use an industrial brand (e.g., Vishay, Panasonic, Wima…). Polystyrene is thermally sensitive, Teflon is mechanically poor, polycarbonate would be fine if it were available. Polyester (e.g., Mylar) will also work fine, though there may be some minuscule measured difference if you live in a high humidity area. If you want to be a contrarian and have the money and space, NP0 ceramic will also perform well.
Thanks, perfect I have some 0.1 WIMA MKP which I can use.
In keeping with what I have, I assume 10uF electro Nichicon KZ at ADJ pin and 120uF electro silmic 2 at output of regulator are acceptable.
Even less critical. I would use at least 105°C rated electrolytics for long term reliability.
Thanks, please – no need to post this comment and clutter up your board – if you could reply via quick email to the address that would be great.
If not, thank you for the help and I assume my question is below irrelevant and shall proceed accordingly.
Q: Trying to use what I have on hand – The 0.1/63V I have are actually FKP, (and not MKP) should be no problem?.
You are correct- no problem. FKP handle higher currents, which isn’t needed here but causes no harm.
I’m a fan of this phono preamp circuit since I first saw it and I couldn’t resist building it with the intention of using it on my EMT 930 turntable (TDS 15 capsule). However, I have been only partially successful. From a practical point of view the preamp works, I mean, it does its job… but the measurements I have obtained are not even comparable to those obtained by Stuart, in fact mine are really bad. Specifically, there is a 1/f noise that is too high to honor its name (His Master’s Noise)… I was able to corroborate this in practice when I plugged it into my system, and my speakers with 100dB of sensitivity didn’t help.
I’ve looked around the circuit a thousand times, I’ve searched for information, but I haven’t found anything that can help me to solve the problem, and the truth is that at this point I’m a bit desperate. Maybe in your blog Mr. Yaniger I can find a solution.
There are no variations from the original circuit, except for adjusting some resistors to get the right values of voltage on the regulators and current on the CSS. A Partidge 977 step-up transformer (6:1) has also been used.
I did some measurements which might show the problem much more clearly (they can be seen at https://www.diyaudio.com/forums/diyaudio-com-articles/163570-masters-noise-thoroughly-modern-tube-phono-preamp-66.html#post6546735)
All help is welcome!
Francisco, a few questions:
When you say 1/f, is that measured at the preamp output, i.e., post-RIAA?
Have you swapped tubes?
Does the noise persist when you short the input?
Stuart, I made measurements at three points, at the SUT output, at the D3a anode and at the preamp output, in all cases with the input shorted. At the output of the SUT I have no 1/f noise, but at the anode of the D3a (using a capacitor to avoid DC) and at the output of the preamplifier, the 1/f noise appears (to give you an idea of the slope, it goes from -115dB at 20kHz to -40dB at 20Hz, in an almost rectilinear pattern). From this I interpret that the problem must be in the first stage. If you tell me how, I could provide you with those graphs, in any case you can see them in the link of my previous post.
As for swapping the valves, the problem is identical on both channels, yet I swapped the valves and obviously there was no change. One could interpret that the problem is the valve, but don’t you think that it is difficult that both are so bad (and so well matched in terms of noise!). Just to be on the safe side I have ordered a couple of extra valves from another supplier, in case that helps.
I also did measurements injecting a signal (1kHz, -6dB) to the input to see the THD, the result was the same noise pattern (although with a higher THD than I would have liked, but that’s another matter…).
The figures you cited, what is the reference for the dB?
I was working with dBFS, mi soundcard is about 2.8V peak before clipping
I’m not sure I submited my previous reply. Sorry if it’s repeated
I did the measurements using the dBFS scale. My soundcard clipping at aprox. 2.8V peak at input.
After reading my previous post, I think it is necessary to clarify that, in the measurements I made on the anode of D3a and the output of the preamplifier, although the noise has the same pattern, obviously the latter is at a higher level (approx. 30dB) as it is affected by the gain of the next stages.
The fact that it’s the same in both channels definitely suggests something common to both. Heater supply might be a place to look first- is it stable and quiet, and is it referenced to AC ground equally on both legs?
One other thing occurs to me- what are the DC cathode and plate voltages?
Sure, I also think it’s a problem common to both channels.
Static measurements right channel (left is similar):
B+ = 249V, B++ = 155V, I (CSS1) = 20.60mA, I (CSS2) = 10.55mA, I (CSS3) = 10.72mA.
Va(D3a) = 114.5V, Vk(D3a) = 1.11V, Va1(ECC88) = 78.2V, Vk1(ECC88) = 1.79V, Vk2(ECC88) = 78.9V
They are not exactly what I expected (especially Va(D3a) which is low because Vk(D3a) is only 1.1V, when the curve in the IR Led datasheet indicated 1.2V for I = 20mA), but I don’t see a serious problem with them, do you?
In reference to the heater circuit, it provides +/-6.1VDC referenced to 60V. This is where I had some doubts in your schematic. Here what I interpreted is that the right channel valves had their heaters connected to the +6.3V branch and the left channel valves at -6.3V (always in reference to a common point raised to 60V with respect to the circuit ground). The last time I measured it was stable but this afternoon I will investigate again if everything is correct. This is a point I hadn’t thought of and it’s worth checking it out.
We can also rule out that the valves are the problem. Yesterday I received a new pair of D3a’s and there was no appreciable change with them in the circuit.
Are the voltages the same in the other channel as well?
One long shot- it’s possible that the batch of IR LEDs is noisy. Take a low voltage electrolytic, maybe 470-1000uF, and tack-solder it across one of the LEDs and see if that has any effect on the noise.
Also, if you’re within a reasonable drive of Phoenix, you’re welcome to bring it by.
Traveling to Phoenix would be fun, it’s only 9000 km from Madrid, Spain 🙂
In any case, thank you very much for the invitation.
Yes, the same for the other channel.
Good idea, I’ll try a capacitor through the IR Led, the truth is that I was a bit surprised about the voltage across it.
Hello SY I recently stumbled across your HMN phono preamp article on diyAudio and really liked it and how readable it was for novices of diy Audio (like me). Given the current global situation i have had a lot of time on my hands and finally decided to restore my dad’s old Dual 704 now running a Shure V15 III with a new stylus. I want to build a phono preamp in the style of HMN for the Shure V15 III and use something like the Sowter 1090 as the input transformer and i saw that you were already working on a MM version of MHN. Have you ever posted an article about it or other material of a MM version? I’d like to read it and see if i can replicate it for the Shure. Thank you for any input you could give.
Hi Tim, this preamp is very much not suitable for MMs like the Shure- the killer is the interaction between an MM’s impedance and an input transformer (and the Miller capacitance of the input stage. I published my MM preamp in Linear Audio, the Equal Opportunity preamp. The performance is excellent, but the downside is that it’s not a simple circuit.
I would like to ask what is considered as ‘acceptable noise level’ for the B Supply of HMN. (<< 5mVp-p?)
Hi Shane, the plate supply is pretty noncritical since the combo of the CCS and the low plate resistance gives us very good power supply rejection. This is especially true at ripple frequencies.
Back of the envelope calculation: the source impedance of the CCS is likely to be significantly higher than 10M at ripple frequencies (Blencowe reports 100M in his excellent book on preamp design). So let’s use 10M as our conservative number. The plate resistance of the first stage (most critical for noise) is likely to be 2k5 or so. So the voltage divider formed by the CCS and plate resistance knocks down any rail noise by a factor of 4000 or so. Probably more. If you have 10mV of ripple (which is a lot), what appears at the plate of the first stage is about 2.5 microvolts. Input-referred, that’s equivalent to about 0.03 microvolts, conservatively. That’s pretty livable! If you use the simple regulators I showed in the article, the ripple is likely to be lower than that, and I was very conservative about the power supply rejection numbers.
Thanks for the comprehensive and helpful reply, SY !
When running the D3a at lower current, I get and LED at 1.5 volts and a current of 12ma for a plate voltage os 140 V.
Does this seem reasonable? I could not find much info on this tube at the lower current.
That seems eminently reasonable!
Thank you Sy!
One more question: To continue using the red LED on the ECC88 at lower current (~5 mA), is reducing power supply from 160 to 140 V the way to go?
You can, but I probably wouldn’t bother- the CCS takes care of setting the plate voltage, and if the B+ is higher, it won’t change the operating point, it will just make the CCS run very slightly warmer.
I found this newer 1.5 Vf, IR emitter, it has the same I-V curve (slope resistance?) as the HLMP-6000:
Vishey P/N TSFF5210-CS12
Gracias, that’s very useful!
dn2540 is unobtainium now, do you know of an available substitute for use in your regulators?
best wishes David
That’s disappointing to hear. Have you looked at Mouser or Digikey to see if they’re carrying another depletion mode poer MOSFET with a high voltage rating?
Well everything critical is out of stock with listed long lead times. But then appears in stock and sells out immediately. Strange times. Well i’ve used the 92 package dn2540’s and they seem okay. Got it all running a few days ago. I salvaged all my SJ74’s from a defunct Denon SACD player which had 24 matched pairs they were all V’s and matched up pretty well! It’s really marvelous sounding. Great design mucho thanks to you and Jan! I had the boards made in india which was a fiasco really. Its more relaxed than the HMN which i have also built and love. I now have to finish my Wright designed RTP-C which is almost finished I just have to sort out the heater supply’s raised voltages. And my go with Morgan Jones Thingy, – and then I’ll have the fully differential set up in place. In the meantime I have an input transformer on my 31 tube SET preamp. I briefly tried the recommended lower current 4ma? for the HMN preamp but will get back to it and see if the same ”relaxed vibe’ of the MM set-up can be achieved. I’ve been using a SC35C with a modern jico line needle and safire atube. Traxks at 1.75 and its all dead quiet and nuanced thanks david
For the HMN, I’d recommend 10mA for the input stage (D3a), then 5mA for the voltage amp and cathode follower (6DJ8/ECC88). The Equal Opportunity is good to go as-is, other than my suggested mod of lifting the input resistor common ground, and returning that junction to ground through a large-ish (maybe 1M) resistor.