Don't know how the 26 compares with the 10Y in a preamp yet, but can easily find out since with a few resistor changes I can use a 10Y in my preamp. I currently have a 30sp (VT-67) in it but I have nice big heatsinks for the DC filament and for the IXYS active loads so I can experiment a bit. I built this preamp as a bit of a test bed. With a more robust power supply I can take the voltage up. Early days yet, but will report!
andy
andy
Kevin,
I know that you have been working with the 26 DHT and line output transformers. I understand that the output level is fairly low due to the output transformer step down ratio. Would it be possible to use a high quality 1:2 input transformer to remedy this, or will this result in overload/clipping since the tube will have to swing more volts?
I know that you have been working with the 26 DHT and line output transformers. I understand that the output level is fairly low due to the output transformer step down ratio. Would it be possible to use a high quality 1:2 input transformer to remedy this, or will this result in overload/clipping since the tube will have to swing more volts?
If I have understood it correctly a 1:1 transformer will give an output impedance of approx the same as the Rp, a high 8k or so. I have a Yamamoto A08 SET, but I feel that a lower output impedance would give me more flexibility/compatibility with other amps, and reduce the interconnect cable influence as well.
Good idea, Kevin!
I am sure that Mr. Eliano or Mr. LeFavre can wind me a suitable transformer when I am ready to take the plunge.
Speaking of DHTs, this might be a good candidate for a parafeed linestage:
EML-20A-20B-30A Datasheet. By Emission Labs.
An externally hosted image should be here but it was not working when we last tested it.
Member
Joined 2006
Just an update on my 26 preamp. I've now fitted IXYS 10M45 active loads (available from Digikey). So setup is 190vDC HT (choke input) into a 150v glow tube (0D3), cascode IXYS active load with a 1.1k sense resistor set for 3mA. Cathode resistor is 2.4k wirewound. The sense resistor for the glow tube is 2.4K, 5watts. Bypass is .056 Russian teflon. Filament supply for the 26 is 6v secondary on transformer, Schottkys into 22,000uF into a LM1084 with a 1.2 ohm sense resistor, 7watts. Heatsink about 3ins square and 2ins high - needs to be a decent size.
Sound is exceptionally clear and nice! Enjoying this preamp a lot. Silvery fast sound with lovely timbre and liquid female vocals.
Not quite sure where I go from here - this is the best yet. I still have to try it with a hotter operating point, and then in goes the majestic 10Y! But for now this is more than good.
andy
Sound is exceptionally clear and nice! Enjoying this preamp a lot. Silvery fast sound with lovely timbre and liquid female vocals.
Not quite sure where I go from here - this is the best yet. I still have to try it with a hotter operating point, and then in goes the majestic 10Y! But for now this is more than good.
andy
That's 10 or 11 mA by my calculations - that's running it pretty hot isn't it!!!
Data says 180v at 6.2 mA
andy
Member
Joined 2006
My 26 preamp is RC coupled, actually built by a friend, and I am thinking of upgrading it to a parafeed type with a IXYS CCS and a matching tranny. And convert the 1086 volt reg to a constant current. Would it work without another voltage regular in front for this filament suppply ?(now it is a resistor and a cap to lower the voltage, thus less heat at 1086).
I understand that the parafeed circuit should be able to separate the signal path from the supply path, and that is great, but what's the benefit of using a matching transformer, suppose just using the capacitor coupling should do the job....say for a short interconnect, no?
And am looking for a tranny for that, how about Lundhal ll1674? does it work here? Or can I go for something more economical like Edcor xsm, or something in between, such as MQ since there is no need to take in DC?
Thanks for the help : )
I understand that the parafeed circuit should be able to separate the signal path from the supply path, and that is great, but what's the benefit of using a matching transformer, suppose just using the capacitor coupling should do the job....say for a short interconnect, no?
And am looking for a tranny for that, how about Lundhal ll1674? does it work here? Or can I go for something more economical like Edcor xsm, or something in between, such as MQ since there is no need to take in DC?
Thanks for the help : )
Kevin's your man for transformer information. But I can confirm that you can use the 1086 as a current source. You don't need to add a voltage reg before it - but use a big cap like 15,000uF. And unless you want a huge heatsink, use the smallest practical secondary voltage on the filament transformer. Nothing over 6v for a 26. I don't know what the threshold is for a 26 but for a 300b at 5v you need about 8v into a LM1084 as current source. So I guess for 1.5v you need 4.5v in minimum - something like that. 6v secondaries give you about 8,5v so you would have headroom for a voltage reg before the current source if you fancied that - won't do any harm!
andy
andy
True. Sound was pretty amazingly energetic and I liked it. I ran it for few hours and turn back to the original 135V/-9A as Kevin suggested me initially. But missing the sound though
The benefit of using a matching transformer is much lower output impedance at the expense of some gain. The mu of a 26 is about 8 and the rp typically around 8K or so depending on operating point. The actual gain you get with rc coupling is highly dependent on the value of the plate load resistor and the external load at the output - in practice you are unlikely to get more than 4X gain or about 12dB.
Using just a 2:1 (recommend >= 10K primary Z, >= 200H primary L) matching transformer with a CCS would actually give you about 12dB of gain into a load impedance of 50K or greater and a source impedance of 2K or so - the lower impedance and 2X the available current implies better ability to drive capacitive audiophile type interconnects and low impedances. MQ makes great parafeed transformers and another strong recommendation for Electra-Print as well.
Should you need much less gain (my case) a 5:1 (15K:600 ohms) gives me 4dB of gain into a >10K load and a source impedance of about 300 ohms which will drive just about anything. Loading into 600 ohms is possible and results in something less than unity gain making the pre just an active buffer. Such a low source impedance makes it possible to use unshielded interconnects without significant noise pick up which is in fact what I am doing. (Relatively inexpensive 1.5M unshielded braided silver wire interconnects not too unlike some of the wildly expensive Kimber cables.)
You can use a pre voltage regulator ahead of the CCS for the 26, and as indicated you need to allow adequate headroom for both the regulator and CCS. Remember that the current setting resistor in both cases drops 1.25V and to allow for this in addition to the voltage required for good regulator performance which I have found to be a minimum of 3V or so to assure that all of the internal current sources are operating in the active region. (Lower than this and the loop gain falls off and performance suffers.) So you should allow for 4.25V per regulator, which would require about 10V raw dc at low line. Each device would dissipate about 3W and the CCS resistor 1.25W (so use at least a 2W @ 70C rated resistor - precision non-inductive ww types are compact and should be suitable) heatsinks with a thermal resistance of 10C/W to ambient of better should be fine. Note that you need a separate filament supply for each 26.
Consider fixed grid bias from a 9V alkaline, eliminates a large bypass electrolytic at the expense of requiring a good small film capacitor at the input. (Much easier than finding a transparent sounding electrolytic.) I'm not a fan of battery bias in the cathode circuit, any NimH sound very inferior IMO to the Sanyo Cadnica NiCd which is much better than anything else I tried, but IMO still far from the last word in resolution or tonal balance.
If you look on my site you will find a link to my several year old 26 transformer coupled pre-amplifier article in Positive Feedback Online.
Last edited:
True. Sound was pretty amazingly energetic and I liked it. I ran it for few hours and turn back to the original 135V/-9A as Kevin suggested me initially. But missing the sound though
Some tubes do sound their best when run at their "upper edge", but given the cost of these tubes anymore, it raises the cost as they will surely not make it as long. If you are willing to pay the cost of more frequent tube changes, go with what moves you. I try to find a happy medium.
Member
Joined 2006
Great, thanks for the info Kevin : )
Can I eliminate the cathode by-pass e-cap anyway since it's a constant current that flows throught the cathode resistor? and can I connect the parafeed cap at the cathode resistor rather than the ground?
And, just one more question, I noticed some filament filter circuits that employ a common mode choke between cathode and regulator, is this something helpful for further removing hum? Any recommendation there?
Thanks so much!
CFT
Can I eliminate the cathode by-pass e-cap anyway since it's a constant current that flows throught the cathode resistor? and can I connect the parafeed cap at the cathode resistor rather than the ground?
And, just one more question, I noticed some filament filter circuits that employ a common mode choke between cathode and regulator, is this something helpful for further removing hum? Any recommendation there?
Thanks so much!
CFT