Of course distortion sky-rockets. The output is no longer a scaled version of the input !
I have a teenager at home....
I try to be so all the time
Would you elaborate on this please..... just a resistor in the collector ?
I found other measures like two diodes between the legs of the LTP but that ruins music.
As we are into it, would you suggest a good Idss compromise for the input jfets that run 1.4mA at iddle
Oh, dear !
I have seen some comment that adding any value of collector resistor ruins the performance of the helper transistor. But equally valid could be the sch that showed a 10k in the collector lead. Both cannot be correct.
Either it ruins the performance, or it doesn't.
that version of offset error correction forces imbalance in the LTP.
Better to find the reason for the imbalance and correct that first. Then you may find that your output offset has magically/scientifically become near zero.
Remember that a jFET that passes a different current from an identical jFET will operate with a different Vgs. That Vgs difference results in an output offset.
Get the two identical jFETs to operate at the same Tj, same Vds, same Id and you will find they have the same Vgs in the quiescent condition. That gives near zero output offset. Nothing left to correct.
BTW,
the same balancing of a BJT input LTP works the same way, with the added proviso that the hFE+Vbe needs to match.
Last edited:
you could use any GR grade of k170.
No one in the UK have bought any of my matched A pairs (of LSK170A). Don't they build jFET LTPs?
I use around 70% to 90% of Idss as my quiescent bias. 5mA Idss set to 4mA (8mA of tail CCS) feeding into a CM gives a maximum output of ~8mApk sourcing/sinking into the VAS helper.
Last edited:
Last EVO.
Higher collector resistor on the VAS that now is a 2SA1318C.
360ohm//2ktrimmer vor bias resistor.... (If the trimmer opens bias goes to 400mA)
Lost the IPS high rail.
FFT is very good and does not change much from 0.48Vdc in to full power.
Left to do: Choose the best Idss for the input jfets
Higher collector resistor on the VAS that now is a 2SA1318C.
360ohm//2ktrimmer vor bias resistor.... (If the trimmer opens bias goes to 400mA)
Lost the IPS high rail.
FFT is very good and does not change much from 0.48Vdc in to full power.
Left to do: Choose the best Idss for the input jfets
Now that changes everything..... Must redo the sim and get back...
Thank you for your support Andrew
PS: Do you still have some of those LSK170A in stock ?
I replied to two other PMs this morning and two more PMs yesterday. So the PM system is working.
I have no PM from RCruz.
I have no PM from RCruz.
I think R cruz is meaning to tell you that he has left a message there in the swap meet. and he wants you to PM him.😉
If that is what happened, then the Forum system has broken again.
I have reported at least twice before that the system loses some of my previously contributed Threads when it does a search. It seems to be a time/duration problem. If a Thread is unused for a while the system does not search and report a new contribution.
I will report again, if that has happened again.
Yes I left a message in the swap meet asking for a PM but now I just sent you a PM Andrew... Please verify your inbox.
I would really like to have some of those matched parts.
I would really like to have some of those matched parts.
A Thermaltrak Badger?
Hi All,
I am about to start my HB, but with a couple of changes as follows:-
1/ 65V rails from a pair of dual secondary 500VA transformers (with properly calculated secondary snubber resistors).
2/ MAT02 super match pair for Q1/Q2 because I happen to have a handful. (the Q2 collector and emitter leads need to be crossed but otherwise fits right in)
3/ MJE15034 +MJE15035 for the drivers for their slightly higher gain and because I have some.
4/ 3 pairs of MJL3281/1302 and, off board, 2 pairs of NJL3281D/1302D mounted on the spare holes at either end of the pre-drilled heatsink. That is NJL-NJL-MJL-MJL-MJL=MJL-MJL-MJL-NJL-NJL with short flying leads back to the PCB and air mounted base and emitter resistors.
5/ A simple string of the 4 thermaltrak diodes with a series resistor to set the bias for the typical operating temperature. This models as horribly over compensated but at least its safe from thermal runaway.
The amplifier will be driving my Manzanitas (the John Busch design). Their minimum impedance is about 7.5 ohm at about 55Hz and about 6 ohm below 20Hz. At the 130V pk-pk output the simulated, instantaneous currents, voltages and phase angles are fairly benign and well inside the SOA curves (derated to 75C). I am happy to have heaps of unused headroom and will also reduce the HB gain.
However my speaker model has not allowed for the power required under load. It only uses the T/S and crossover values.
This leads to a couple of questions, if I may...
1/ Has anybody successfully mixed standard MJL's with the Thermaltrak NJL? They're the same transistor die right? I have thought to increase the value of the emitter resistors to 0R68 to guarantee load sharing, is this a good idea?
2/ What is the effect of a real load on a speaker? Does it simply increase the resistive part of the apparent motor load or are there more complex issues involved?
Season's Greatings to all and special thanks to OS and the diyTeam for a great little design.
John
Hi All,
I am about to start my HB, but with a couple of changes as follows:-
1/ 65V rails from a pair of dual secondary 500VA transformers (with properly calculated secondary snubber resistors).
2/ MAT02 super match pair for Q1/Q2 because I happen to have a handful. (the Q2 collector and emitter leads need to be crossed but otherwise fits right in)
3/ MJE15034 +MJE15035 for the drivers for their slightly higher gain and because I have some.
4/ 3 pairs of MJL3281/1302 and, off board, 2 pairs of NJL3281D/1302D mounted on the spare holes at either end of the pre-drilled heatsink. That is NJL-NJL-MJL-MJL-MJL=MJL-MJL-MJL-NJL-NJL with short flying leads back to the PCB and air mounted base and emitter resistors.
5/ A simple string of the 4 thermaltrak diodes with a series resistor to set the bias for the typical operating temperature. This models as horribly over compensated but at least its safe from thermal runaway.
The amplifier will be driving my Manzanitas (the John Busch design). Their minimum impedance is about 7.5 ohm at about 55Hz and about 6 ohm below 20Hz. At the 130V pk-pk output the simulated, instantaneous currents, voltages and phase angles are fairly benign and well inside the SOA curves (derated to 75C). I am happy to have heaps of unused headroom and will also reduce the HB gain.
However my speaker model has not allowed for the power required under load. It only uses the T/S and crossover values.
This leads to a couple of questions, if I may...
1/ Has anybody successfully mixed standard MJL's with the Thermaltrak NJL? They're the same transistor die right? I have thought to increase the value of the emitter resistors to 0R68 to guarantee load sharing, is this a good idea?
2/ What is the effect of a real load on a speaker? Does it simply increase the resistive part of the apparent motor load or are there more complex issues involved?
Season's Greatings to all and special thanks to OS and the diyTeam for a great little design.
John
New build help
I've built my first badger and i'm having an issue with my readings from TP1 to TP2, they are too high at 0.250-0.280vdc, i understand they should be around 0.025vdc.
Reading through this forum I've found these values might be useful:
Measure across R14 is 8.2vdc.
Measure across Fuse1 and Fuse2 is 0.022vdc each.
I have it powered by the DIYAudio PSU and an Antek 8445.
Anyone have an idea of where it might be broken or what i can probe to get a better idea?
Thanks
I've built my first badger and i'm having an issue with my readings from TP1 to TP2, they are too high at 0.250-0.280vdc, i understand they should be around 0.025vdc.
Reading through this forum I've found these values might be useful:
Measure across R14 is 8.2vdc.
Measure across Fuse1 and Fuse2 is 0.022vdc each.
I have it powered by the DIYAudio PSU and an Antek 8445.
Anyone have an idea of where it might be broken or what i can probe to get a better idea?
Thanks
Hi,
today my v. 2.4 PCBs successfully arrived. Very fine craftmanship, indeed!
Before I start collecting the components and populating the boards, I'd wish to know what ist the latest version of this amp. As questioned here, there might be some incongruencies between the build guide and some decent version, discussed elsewhere. Who would like to help by answering these questions?
Best regards!
today my v. 2.4 PCBs successfully arrived. Very fine craftmanship, indeed!
Before I start collecting the components and populating the boards, I'd wish to know what ist the latest version of this amp. As questioned here, there might be some incongruencies between the build guide and some decent version, discussed elsewhere. Who would like to help by answering these questions?
Best regards!
Have one question not necessarily related to HB.
Is it worth PAYING (and using) using SANKEN 2sa1295/2sc3264 transistors, which cost twice as much as somewhat less powerful sanken alternatives: 2sa1294/2sc3263 ?
As drivers in my surround active HB system I plan to use sanken 2sa1859a/2sc4883a. Or will 2sa1837/2sc4793s perform better in HB?
Thank you,
cheers,
Is it worth PAYING (and using) using SANKEN 2sa1295/2sc3264 transistors, which cost twice as much as somewhat less powerful sanken alternatives: 2sa1294/2sc3263 ?
As drivers in my surround active HB system I plan to use sanken 2sa1859a/2sc4883a. Or will 2sa1837/2sc4793s perform better in HB?
Thank you,
cheers,
Hi Kay
- R3, R6 = 22 kΩ in the V2.4 schematic vs. R3, R6 = 33 kΩ in the build guide is to reduce the overall amplifier gain from about 40 to 26
- R23 = 47 Ω in the V2.4 schematic vs. R23 = 33 Ω in the build guide for the higher CCS current
- R24 = 1.5 kΩ in the V2.4 schematic vs. R24 = 820 Ω in the build guide, gain related.
- R26 = 10 Ω in the V2.4 schematic vs. R26 = 100 Ω in the build guide, higher CCS current
- R27 = 82 Ω in the V2.4 schematic vs. R27 = 68 Ω in the build guide, increases the current to about 10mA and helps remove ringing for square wave or clipping
- the "diode fix" (baker clamp or D-BC) also removes output ringing - BAV21
- the LC option - add a 2pF cap across R6 to slightly improve HF response
John