How many times I’ve been asked “how do I lower the gain of the DHT preamp?” many builders love the DHT sound and yet don’t really need that gain. Of course when you use a transformer load, step down is the obvious choice. And it works really well when you invest in a very good quality OPT.
However, when you use DHT with high rp, generally most of the low current ones, you face a tough challenge with regard to frequency response.
With the hybrid mu-follower topology (aka gyrator) we have the problem that the gain is fixed to nearly mu.
I’m in love with the DHT sound and in a preamp a 01a and the 801a/VT-25 DHT Preamp Update preamp cannot be beaten in my view. They all deliver a gain of 8 with a hybrid mu-follower topology.
This can be overcome in the following way:
We add an attenuator network at the mu output before the source follower circuit. In the above example, I configured the resistor divider R2 and R3 to provide 1/4 gain so the resulting gain is around 2 (6dB). This works like a charm and helps reducing the microphonic noise due to the attenuation of the overall signal. However, the signal headroom is limited but this shouldn’t be a problem as we would expect to run this below 5Vpp.
The rest of the circuit remains the same. The gyrator is the Rev08 and set to about 90-100V depending on the valve, the source follower is configured to operate at 15mA and the cathode SiC bias is using the C3D02060F which are in my experience the best sounding arrangement for this preamp. The filament regulators are the classic Rod Coleman regs.
It can be improved further by adding a DC level shifter to the bottom of R3 (a simple PMOS would work here like shown here) to increase the input DC level of the Source Follower. This will optimise the performance and reduce even further the distortion at higher voltage swing. Of course, this adds more complexity to the circuit.
I will likely hack my 01a preamp as don’t need the gain of 8. A simple gain of 2 would do just fine in my current 300B amplifier.
Great DHT 01 and very interesting design.
I will test this in my CX 301A preamp.
I did somesing wrong with vote!
Hop you can change to 5 stars!
Best regards
AP
Thank you Alf. Please test and report your feedback.
Don’t worry about the starring, I don’t mind. It got down to 3.7 I think 🙂
Hi Ale, very interesting post. The high gain has always put me off using the mu-follower arrangement. Would another possible option be a transformer volume control at the output rather than a conventional pot on the grid? I’d like to avoid the added complexity of the source follower if possible. Many thanks for your inspiring work!
Hi Paul, yes you can. However, I’d use the AVC/TVC at the input if the signal level is too high for it, it may increase distortion. You need to try and listen to define where it works better for you. Mind you that if the amplifier’s input impedance is low (normally if you have a solid state amp), you will need a source follower to avoid slew rate distortion and loss of high frequency response. You can build the pre-amp and leave space and provision for a Source Follower. Try it without it and if you don’t need, you don’t add it to the preamp. Otherwise, am afraid this is a limitation that all low-current DHTs have. There is no free lunch!
Nice !
The 01a preamp I’ve been waiting for 🙂
Hi Ale,
I see you are running 01a at 4-5mA, is this ok? I assume this will reduce distortion. I am running SF boards at 20mA, I could probably get away with15mA or even 10 maybe.
Hi Barry, yes. I run them somewhere between 3 and 5mA. It depends on the state of the valve as I look to bias them about 90-110V. Higher current will burn emission faster, but on the other hand you get (slightly) lower distortion and better current drive. No need to push for higher current in my view. I’d rather add the Source Follower to provide the right current drive to the amp, etc. The Source follower can work at 10mA for sure. I found that distortion is lower (higher odd harmonics in particular) when current is higher for obvious reasons as transconductance of the MOSFET increases so does the feedback of the follower topology which linearises the transistor more.
Hi Ale! I was curious about this project and I finally managed to almost finish it. It plays good, however it could play better due to correct gyrator settings. Maybe you could help ?
I have a problem with the anode voltage – instead of 90-100 volts, I have 78 volts at a current of 5 mA. How to adjust the gyrator to obtain a voltage of about 100 V on the anode without exceeding 5 mA of current?
Gyrator parameters:
R4=220 K
R6=4.7-10 M
M3=DN2540
J1-J4:BSH111BK
R7=470 Ohm
Thank you in advance.
Best regards.
Hi Paul, what do you mean by “it could play better due to correct gyrator settings”? It looks to me that it’s working well. The valve is the one to point out at. As an old device and parameter variance you are not able to dial same operating point. Try a different set of 01a or if you have implemented the same circuit as on this blog post, you may try adding a bit of series resistance to the array of SiC diodes to move the bias point. Try adding something like 100ohms which should move the bias point about 500mV. Then you can readjust the gyrator setting which will modify “only” the anode voltage.
5mA is good enough current in the 01a at that lower anode voltage anyhow.
Cheers, Ale
Hi Ale, I mean, this preamp sounds amazing. The bass is great, the dynamics are superb with my DSD R2R DAC, and the vocals sound lifelike. However, I believe mine could perform even better with a higher voltage on the plate. Thanks a lot for your help. I’ll check this cathode resistor.
Cheers,
Paul