UX-201-A, CX-301-A tubes have remained my preference for input stage for a long time now, hybrid mu-follower and filament biased, yes. For 15 years now 🙂
Actually, I have one double stamped version with both types and respective logo's on either side of the single base (one tube). I agree that its time to push the price of these tubes up to where they should be.
Theres a lot of other good stuff about this triode which has gone so far unmentioned - its worth looking into.
I also agree with ~-4.5Vg, 3 - 3.5mA Ip, ~100Vp but with high AC and low DC grid impedance to signal common - even if the tube is ok with 1M 'grid leak' resistor (low grid current type, get away with minimal biasing etc)
LH/S
Actually, I have one double stamped version with both types and respective logo's on either side of the single base (one tube). I agree that its time to push the price of these tubes up to where they should be.
Theres a lot of other good stuff about this triode which has gone so far unmentioned - its worth looking into.
I also agree with ~-4.5Vg, 3 - 3.5mA Ip, ~100Vp but with high AC and low DC grid impedance to signal common - even if the tube is ok with 1M 'grid leak' resistor (low grid current type, get away with minimal biasing etc)
LH/S
Last edited:
Since ~30´s mica was on rise, also vacuum pumps technology were better. Maybe the vintage sound is, because of low quality vacuum-more ionic currents
Hi Bela,
At the low current involved in this circuit (i.e. 3mA) you can use a low capacitance JFET like the 2SK170 which helps improving significantly the HF response. In a cascoded pair when VDS<2V due to the top FET bias requirements, the DN2540 capacitances rocket up massively whilst the 2SK170 for example has only Ciss=50pF max at VDS=1V.
And it sounds beautifully!
Ale
Recently I use DN2540-IXTP01N100 (1000V, 120pF Ciss) cascode current sources.
I lost half dozen 2540s (over 400V B+) in my tube measuring set, due to the voltage spikes ( protector zeners also destroyed).
1kV DFET is more stable..... or typical survivor? 🙂
I lost half dozen 2540s (over 400V B+) in my tube measuring set, due to the voltage spikes ( protector zeners also destroyed).
1kV DFET is more stable..... or typical survivor? 🙂
Yes, same experience here. When using a CCS for a HV shunt regulator I had several ocassions the DN2540 dying as well. I ended up replacing them for IXTP01N100D as well but had some limitations at setting high currents e.g. 90mA in a cascoded arrangement. Other than that, the IXTP01N100D is rock solid.
Any reason why the smaller Jfet could not pass higher current levels, say 20-22mA? The 2sk170 have a V grade that go close to the that much current at Idss, and if this is inadequate, you could look to Jfets like the j309/10. Dissipation should not be an issue if I have figured correctly. Just wanting to double check.
Applied just the J310 in the circuit as proposed by Ale (cascoded gyrator load) to run a 4P1L at 30mA. Works just fine.
Ulrich
Not really, could be many things but didn't bother as below 100dB in my test bench could be anything. Either way, it sounds nice!
Ale, I've been following your 01a preamp adventures with the mu-follower output. In one of your blog entries you mention using it since it offers plenty of gain.
Is it still a good choice if we only need unity gain, or would transformer coupled be ok (ie similar slew rate, etc)? Thanks.
Is it still a good choice if we only need unity gain, or would transformer coupled be ok (ie similar slew rate, etc)? Thanks.
#01a output impedance is on the high side, at 3mA anode current about 10k, even at 6-10mA region about 7k.
Good 15k:600 transformer's cost is at least as large as the whole #01a preamp with active load!
The output impedance of a hybrid mu follower is 1/GM or about 50r if you're using the 2SK170 as lower FET. If you need unity gain then either use a pot or a different circuit..
I am using LL2745 transformers with a 4mA gap (ALT-R for low impedance but also low output). They're about $140 each. If you have Interconnect cables longer than 3 feet, that could be a choice. However, It was before Ale release his new circuit and board.
Looks like this
I use AVC at output and it's dead quiet. The filaments are a bit starved (0.230A). As you can see, Rod regs are there and Tom Cr 21thMaida reg also. Tunings are therefore easier when I'm rolling tubes. CX-301A and UX-201A require different settings (01A on picture).
I use AVC at output and it's dead quiet. The filaments are a bit starved (0.230A). As you can see, Rod regs are there and Tom Cr 21thMaida reg also. Tunings are therefore easier when I'm rolling tubes. CX-301A and UX-201A require different settings (01A on picture).
Attachments
Nice built Crazyfrog. Are you using Slagle AVCs?
Have you compared the sound to standard pots ?
Best,
Radu
Have you compared the sound to standard pots ?
Best,
Radu
Thanks. Yes they come from Intact audio. I tried both stepped attenuator and AVC on a 10Y line amp (LL1660 at output) but not on 01A pre. However the results were good enough to try it on the actual unit and I'm not disappointed. Full signal goes at the triode grid in that case, that must not be bad!
Thanks guys. I am attracted to the idea of the hybrid mu-follower since it removes the OPT from the signal path... but at the same time I like transformer coupled SET amps I've built 🙂 At this level it all ends up expensive in the end so OPT vs none for costs isn't a big deal.
If I go with the hybrid mu-follower I'll use a voltage divider at the input. My setup is a bit unusual since I am using it in a multi-channel digital only setup, that's why I don't need any gain. Minor variations in gain can be accommodated in the convolution filters.
If I go with the hybrid mu-follower I'll use a voltage divider at the input. My setup is a bit unusual since I am using it in a multi-channel digital only setup, that's why I don't need any gain. Minor variations in gain can be accommodated in the convolution filters.