I have tried using two Hammond 156C in series as plate chokes for a 26 tube. They sounded better to me than IXYS active loads, but were rather prone to hum and hard to get quiet.
What would be the ideal plate choke for a 26 with its Rp of around 8.5K?
Looking at Lundahl
LL1667-05mA 810H
LL1667-07mA 580H
LL1667-10mA 405H
Which of those?
Any experiences with these, or something else again?
Andy
What would be the ideal plate choke for a 26 with its Rp of around 8.5K?
Looking at Lundahl
LL1667-05mA 810H
LL1667-07mA 580H
LL1667-10mA 405H
Which of those?
Any experiences with these, or something else again?
Andy
Perhaps the Hashimoto HL-10 K-6 is an alternative:
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
DJN:
The choke is in lieu of a plate resistor or SS load. The choke offers the advantage of very low resistance, compared to a resistive load, so low B+ can be used. The choke must be closely matched to the tube, with enough inductance to provide good low frequency response.
At any given frequency, the load, as seen by the tube (plate load only, now) is 2*PI*L*F=the inductive reactance, or AC resistance where L is measured in Henries, F is the frequency. Add this number to the DC resistance (DCR) and you get the load on the plate.
To a point, the higher the load resistance, the greater the stage gain. Obviously, the lower the frequency, the lower the "resistance" hence lower gain. To keep from causing a low frequency roll off, enough inductance (L in the equation) must be present for the desired bass response.
Just another way of designing a gain stage. Simple but very expensive.
Stuart
The choke is in lieu of a plate resistor or SS load. The choke offers the advantage of very low resistance, compared to a resistive load, so low B+ can be used. The choke must be closely matched to the tube, with enough inductance to provide good low frequency response.
At any given frequency, the load, as seen by the tube (plate load only, now) is 2*PI*L*F=the inductive reactance, or AC resistance where L is measured in Henries, F is the frequency. Add this number to the DC resistance (DCR) and you get the load on the plate.
To a point, the higher the load resistance, the greater the stage gain. Obviously, the lower the frequency, the lower the "resistance" hence lower gain. To keep from causing a low frequency roll off, enough inductance (L in the equation) must be present for the desired bass response.
Just another way of designing a gain stage. Simple but very expensive.
Stuart
sepolansky's answer is good. For a little more, choke loads compared to other loads here: http://www.ecpaudio.com/pdf/parafeed_basics.pdf
In driver applications the choke load also allows you an output swing of almost twice the dc supply rail by virtue of the energy stored in the magnetic field of the choke. This is quite convenient for driving low mu dhts for example on relatively driver stage plate voltages. It also means that for a given swing there is the possibility of slightly better linearity provided that the stage is loaded by a high impedance load.
FWIW I'd use the choke rated for 7 or 10mA as I run 26s in my line stage at 6mA and above. Note also that the plate to line transformer suggestion is a very good one. (This is what I have been running for over 10yrs now in my 26 based line stage.)
Any choke is going to have pick up issues with external AC magnetic fields, this is also true with transformers. Core geometry may help a little, care in physical placement and shielding can help a lot.
FWIW I'd use the choke rated for 7 or 10mA as I run 26s in my line stage at 6mA and above. Note also that the plate to line transformer suggestion is a very good one. (This is what I have been running for over 10yrs now in my 26 based line stage.)
Any choke is going to have pick up issues with external AC magnetic fields, this is also true with transformers. Core geometry may help a little, care in physical placement and shielding can help a lot.
Last edited:
Anybody seen or used this guy's plate chokes?
ANODE GRID CHOKE 40H 200H 300H 500H 12AX7 12AU7 6SN7 83 on eBay (end time 02-Mar-11 05:36:18 GMT)
ANODE GRID CHOKE 40H 200H 300H 500H 12AX7 12AU7 6SN7 83 on eBay (end time 02-Mar-11 05:36:18 GMT)
Andy,
I’ve had good luck with Magnequest EXO-010 Plate chokes (10ma/275H, 825 DCR), 04-006 gapped to 15ma/200H, EXO-99 (10ma/750H, 2200 DCR)--prefer the EXO-010s. I’ve also tried Electroprint AC9037 (10ma/100H) but I think the inductance was not sufficient. All are used with parallel feed preamp outputs, QEE-3436 (series primary 20K/600) and Magnequest B7-15K.
Matt
I’ve had good luck with Magnequest EXO-010 Plate chokes (10ma/275H, 825 DCR), 04-006 gapped to 15ma/200H, EXO-99 (10ma/750H, 2200 DCR)--prefer the EXO-010s. I’ve also tried Electroprint AC9037 (10ma/100H) but I think the inductance was not sufficient. All are used with parallel feed preamp outputs, QEE-3436 (series primary 20K/600) and Magnequest B7-15K.
Matt
I'd agree, about the minimum you would want here would be 200H..
Andy, I'd be very wary of buying a plate choke from anyone who does not also wind interstage or output transformers as some care in winding and design is required to keep the shunt and stray capacitances manageable, particularly as the inductance goes up. I'd go for Lundahl, Sowter, Electra-Print or Magnequest chokes new or used in preference to the eBay ones you ask about. Even great chokes often roll off surprisingly early on the HF end IMLE..
Last edited:
Hi Andy,
I'm actually mildly notorious as a bolster of plate to line transformer topology in 26 applications. I'm using vintage UTC HA-133 permalloy transformers in my line stage, and always have.
I think any of the three listed Lundahl chokes is going to provide reasonable performance in your application as long as you keep the choke clear of external magnetic fields, and the input impedance of the power amplifier is sufficiently high as not to constitute a significant load on the 26. Note that source impedance in this configuration is a very high 8K or so making it susceptible to external electro-static coupling.
Assuming that the usual trade offs apply with chokes in that the higher inductance versions probably have significantly greater shunt capacitance (limiting HF bandwidth) I would then choose the smallest value that would provide the LF extension you are looking for.
Or I would go for the transformer... Much lower output impedance, and a small amount of gain suitable for most line stage applications. I wrote an article for Positive Feedback Online about one of my 26 DHT line stage designs some years ago. You can find it here: http://www.positive-feedback.com/Issue16/kennedy.htm Might provide you with some useful insights. Hope you are using Rod Coleman's dht filament ccs in your project..
Last edited:
Hi vgeorge,
Hmmmm - so the LL1667 can be improved on. What sort of capacitor were you using with it? I could try out the primary of the LL1660/5mA as a plate choke and see what that sounds like.
Were you using a PP circuit or was the LL1660/PP pretty much gapped for 5mA for SE use?
Reason I'm thinking plate choke for the 26 is I want to direct couple to the next stage.
andy
Hmmmm - so the LL1667 can be improved on. What sort of capacitor were you using with it? I could try out the primary of the LL1660/5mA as a plate choke and see what that sounds like.
Were you using a PP circuit or was the LL1660/PP pretty much gapped for 5mA for SE use?
Reason I'm thinking plate choke for the 26 is I want to direct couple to the next stage.
andy
Well - I was inspired to put together a 26 preamp by vgeorge!
I used filament bias - 10 ohm resistor. And hooked up my LL1660/5mA. 145v on the plate of the 26, 10v on the top of the cathode resistor, 5mA through the 26.
Well, this clearly works very nicely. Pleasant musical sound. I could live with this. I haven't optimised the HT or the filament supply yet - just running that off a bench supply.
It doesn't quite have the spooky presence of the 10Y into the Hammond 126C, but that's with optimised HT and filament supply, so there must be more to come. This is a very good start. I'll listen to this for a while.
Andy
I used filament bias - 10 ohm resistor. And hooked up my LL1660/5mA. 145v on the plate of the 26, 10v on the top of the cathode resistor, 5mA through the 26.
Well, this clearly works very nicely. Pleasant musical sound. I could live with this. I haven't optimised the HT or the filament supply yet - just running that off a bench supply.
It doesn't quite have the spooky presence of the 10Y into the Hammond 126C, but that's with optimised HT and filament supply, so there must be more to come. This is a very good start. I'll listen to this for a while.
Andy
I used an 0.470nF FT-3 russian teflon.
I am using the 1660 for SE use, 133v B+ and a 9v battery.
What is your voltage on the HT supply for the fillament bias?
I am using the 1660 for SE use, 133v B+ and a 9v battery.
What is your voltage on the HT supply for the fillament bias?
Hi Kevin,
i read your interesting article at Positive Feedback Online about your 26 Pre-Amp.
Did you have any concerns about using the Hashimoto HL-10 K-6 instead of the UTC HA-133?
Also the Tentlabs DHT Supply should work proper with the 26, right ?
I have 145v on the plate of the 26, so HT supply into the LL1660/5A will be just a little bit above that - I didn't measure it. The dropper resistor at present from my 330v HT supply is 19,600 ohms which would indicate about 5mA going through each 26. That's a pretty similar operating point to what you'd use with a 9v battery. The 10 ohm cathode resistor dictates approx. 10v at the cathode, though this varies with the age of the 26s used and their filament consumption.
I obviously don't need 330v to start with so that needs a re-design. I fancy choke input with a nice AZ1 mesh or an 80.
vgeorge - have you tried filament bias? It's worth a try - it can be really good. Your HT would stay the same. You'd just need a 10 ohms cathode resistor rated around 50 watts and a filament supply that would produce around 11 to 11.5v at 1 amp between one side of the filament (other side from the resistor) and ground. The filament supply has to be very pure to get the best out of it.
Andy
Matched to the tube? Why?
An inductor has a DC resistance and an AC reactance. In order for it to work as a plate load, it needs to be operating in its inductive region. I.e. XL = 2*pi*f*L needs to be significantly greater than the DC resistance. I'd use XL >= 10*DCR as a starting point.
I picked up a pair of Electra-Print 100 H inductors. They have a DCR of about 1400 ohm. I'd expect them to be good down to (or slightly past) 1400*10/2*pi*100 = 22 Hz.
PSpice says the gain of the input stage is 0.3 dB down from its mid-band value at 22 Hz; -3 dB at 5 Hz. That's also what I seem to recall measuring in the lab.
Actually that isn't correct. The reactance and resistance are 90 degrees out of phase. The total impedance magnitude seen by the plate is |ZL| = sqrt(DCR^2+XL^2).
~Tom
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.