Unfortunately although you are operating in constant current mode removing that bypass capacitor will result in a much higher rp which will result in rather poor HF performance.
I would recommend trying battery bias as previously discussed using RC coupling in the grid circuit to inject the bias voltage - this would allow you to remove both the cathode bias resistor and the bypass capacitor.
The parafeed cap can be connected to the cathode connection in any of the situations described above, however the cathode bypass cap is required for the reasons mentioned.
Incidentally the cathode bias wants to set your operating point as well as your CCS, in this case you should allow the CCS to do it. If you do not want to use the battery bias mentioned something like 5 red leds in series in place of the cathode bias resistor & cap should make for an audible improvement. You can try connecting the parafeed cap between the leds and the filament. (Try both there and normal circuit ground - you may or may not prefer one over the other.
Hope this all helps.
Hi,
How long the tube will survive if I continue the same op points? At least for an hour daily, I am using this Pre nowadays. I don't mind of changing tubes every 2 years and having the same sound which I like.
Probably OK if you have a target life span of a couple of thousand hours or less.
Member
Joined 2006
Hi Kevin,
Just saw your post on Tubes to transistors : )
Could you share your thoughts or experience on using something like a FW B1 buffer placed directly at the coupling output cap? Given that the buffer is said to be quite neutral, would you think that the output transformer can be replaced?
Thanks!
Just saw your post on Tubes to transistors : )
Could you share your thoughts or experience on using something like a FW B1 buffer placed directly at the coupling output cap? Given that the buffer is said to be quite neutral, would you think that the output transformer can be replaced?
Thanks!
Hi Kevin,
Tried 150v/6ma and sound is pretty impressive. I think I can settle down with this as well. Bass response is slightly inferior but overall sound is perfectly fine. Thanks.
I've never heard a B1, but the word on the street is that it is quite neutral - if you have one on hand it would be a worthwhile experiment. I am not sure what the input impedance of a B1 might be, but the higher the better from the perspective of cap size, and load on the 26..
Last edited:
Kevin refers to red LEDs, and I just tried those in place of a cathode resistor. Initially I was excited - very clear and dynamic. But there was an edge to the sound I couldn't live with, and I took them out and put in a wirewound cathode resistor. I found that fine - the smoothness returned and sound was only marginally less clear and dynamic.
I'm using cap out, and I didn't find any loss of treble without a bypass cap so I left it out, so with cap coupling that's a good option. With transformer OPT, just as Kevin says, you want to keep the Rp low.
As for operatint point, I've been using a 150v glow tube and that's really too low - I need two to give 210v or 265v depending on the tubes. So with 3mA through the 26, the sound really is a bit thin. So will need to up the HT and the current as said in previuous posts. Looks like the 26 likes to be run hard - or at least quite hard!
andy
I'm using cap out, and I didn't find any loss of treble without a bypass cap so I left it out, so with cap coupling that's a good option. With transformer OPT, just as Kevin says, you want to keep the Rp low.
As for operatint point, I've been using a 150v glow tube and that's really too low - I need two to give 210v or 265v depending on the tubes. So with 3mA through the 26, the sound really is a bit thin. So will need to up the HT and the current as said in previuous posts. Looks like the 26 likes to be run hard - or at least quite hard!
andy
Hi Andy,
Actually I tried all these biasing methods and settle down with battery bias as Kevin suggested because it gave the most natural sound out of all. It reduced the hum considerably too. About the op points; if you playing Rock music, run those 26 hard as you can. It will bring the band to your house
. Also it will give life to vocal jazz too. This is what I observed so far.
Hi Kevin,
Tried 150v/6ma and sound is pretty impressive. I think I can settle down with this as well. Bass response is slightly inferior but overall sound is perfectly fine. Thanks.
This is where I run mine, and is the maximum plate current I can run without danger of saturating the UTC HA-133 plate to line transformers. It's pretty authoritative, particularly when I am not overtaxing my 8Wpc 300B SE amplifier on loud, low pitch music. The new GM70 amp should resolve that issue if I ever get it done.
This is very like the reaction I felt to using red-LED biased current source in the shunt-cascode driver for my 300B-SE. I took it out again and used cross-coupled transistor current source.
Curious
Ring of two type CCS? This is my general preference as well for bipolar transistors. In general I have avoided CCS in the audio path of my own equipment. I am experimenting a little with some of the IXYs devices - SY, Doug and a couple of others think very highly of these which has my curiousity piqued.
I was thinking perhaps I had a specific preference for bias approach but I think I might be almost close to agnostic. I use cathode bias with tubes like the 6J5/6SN7/6SL7 almost exclusively, fixed (grid) bias in power stages and with DHT types in general or and some high transconductance types like the 5842. I have also employed LED bias with 5842 and D3A with choke loads or tube based gyrators. I've found the LED bias sounds good in these applications sans CCS. Using high transconductance types with LED bias in particular means that you need to pay close attention to matching and target plate current vs supply voltage (since I don't currently use CCS).
Fixed bias seems to work really well with low transconductance dhts and results in reasonably predictable Ip which generally is not the case with higher transconductance types which need to be tweaked to achieve the desired operating point.
I guess there are about as many ways to do this as there are designers.
Hi Andy,
Like most of my system I designed and constructed the speakers myself - it was, so to speak "the final frontier"
They are Onkens which are based on the old Jensen Ultraflex design. Mine use Iconic 165-8G 16" HE woofers bought new (basically the same as an Altec 515-8G) and vintage pro JBL 4333 mid horns with acoustic lenses/diffusers on them and JBL 2402 annular ring radiators. These are all tied together with an external 2nd order Butterworth x-o using generic Madisound air core inductors and Clarity SA film caps. You can find more details on several of the older Onken threads on our speaker forum and some pictures in my profile. They are efficient - the bass bins just a tad under 100dBspl@ 1W/1M and the mid/high horns are over 110dBspl @ 1W/1M allowing the use of a much smaller amplifier if I decide to bi-amp - who knows bass by GM70, highs by 300B? It's just too much fun!Edit: I didn't mention that the speakers were designed for use with SE amplifiers with a relatively high output impedance in the range of 2 - 3 ohms. This includes the cabinet design & bass alignment and the X-O electrical design as well. It works quite well I am happy to report, but it is the second X-O - the first one was marred by a few errors in the design analysis.
FR wise in room they are better than +/- 2dB from 100Hz - 10kHz, and are down 3dB around 35Hz, and 16kHz. (the annular ring radiator is great to 15kHz and starts to roll off at that point) I used Arta and/or audiotester to measure in room response with a once calibrated behringer mic.
Last edited:
Hi Andy, have you tried a para-feed solution in W.E. style? The only difference would be in the CCS, solid state in your case vs the original inductor.
You could use a relatively cheap and very good Sowter:
Sowter Type 3232
To complete the design you could go for the Sowter transformer attenuator at the input.
Cheers,
45
You could use a relatively cheap and very good Sowter:
Sowter Type 3232
To complete the design you could go for the Sowter transformer attenuator at the input.
Cheers,
45
Hi 45,
That's an option - thanks for that! I could try parafeed with a Lundahl OPT just to see. I could try choke loading or active load - did you find a preference if you tried both? Can you remind me of the WE setup with a SE stage? I'm thinking of the PP version of a cap between shared cathode and HT of transformer.
andy
That's an option - thanks for that! I could try parafeed with a Lundahl OPT just to see. I could try choke loading or active load - did you find a preference if you tried both? Can you remind me of the WE setup with a SE stage? I'm thinking of the PP version of a cap between shared cathode and HT of transformer.
andy
It is the driver stage. More precisely the para-feed driver of the WE 92a, for example:
Western Electric - Rosetta Stone for Triodes
You just put a CCs (or a choke) in place of the 100K resistor and play a little bit with the caps. However I think you need the input trans to get the best out of it.
This is a recent one (look at the input stage): http://www.nutshellhifi.com/GarysAurora.gif
Another interesting schematic is that of the Klimo Beltaine. The output stage uses a SE 300B through 1/2 primary of a PP output trans, the DC induction is balanced using an EL 34 connnected on the other half. You could use the same principle for the preamp.
I don't have schematic right now but I have seen it somewhere. This little 5W amp was really delicious. It is a pity it has been discontinued.....
Cheers,
45
Last edited:
Hot!
There is one way of biasing a DHT not mentioned here. When using the heatercurrent to get the necessary cathode biasvoltage no decoupling cap will be needed.
This is just an HOT idea, have not tried it myself yet. The cathode-resistor will dissipate over 8W! Think I have seen it somewhere before and I´m going to use it on a small 3A5 parafeed.
There is one way of biasing a DHT not mentioned here. When using the heatercurrent to get the necessary cathode biasvoltage no decoupling cap will be needed.
This is just an HOT idea, have not tried it myself yet. The cathode-resistor will dissipate over 8W! Think I have seen it somewhere before and I´m going to use it on a small 3A5 parafeed.
Attachments
