Ok that one is most likely in need of a buffer. Try making a twisted BOSOZ or just the plain vanillia original BOSOZ. It should solve that problem, as I have never had problems driving the amps I've made so far with such (I have never made the twisted BOSOZ, ony heard it a Steenoe's place)
Magura
Magura
Actually, the previous schematic is my little single-ended in to balanced out converter. It's driving my little ZV7-T-J (jfet) experiment as I write this. I just happen to like the buffer at the front end of my power amplifiers so that anything can drive them. I also like the delicate high notes the buffer produces. I don't know if the bandwidth is expanded, but it sure sounds like it is.
Ever use this buffer?
I have the "twisted" kit. I've just been too busy learning from my projects to construct it. Someday...
John
Ever use this buffer?
I have the "twisted" kit. I've just been too busy learning from my projects to construct it. Someday...
John
Magura, When I experimented with <ZV4 the feedback Rs and such really limited the high end. When the ZV4 buffer came along, it really solved the problem pretty well without to much added circuitry, pure Zen style
When I experimented with the JFET ZV8 & 9 ideas, it reminded me a little of the progression to ZEN4. I wanted a little more gain than the single part was able to do and enough input Z for anything to drive it. Also, the high frequency response was a little lacking without low value resistors. Adding a parallel JFET would only add more capacitance to the high freq problem... So, I'm going to go ahead and try 2 parallel devices but, driven with the low buffer source impeadance and isolated from the feedback network resistors... Like ZV4
When I experimented with the JFET ZV8 & 9 ideas, it reminded me a little of the progression to ZEN4. I wanted a little more gain than the single part was able to do and enough input Z for anything to drive it. Also, the high frequency response was a little lacking without low value resistors. Adding a parallel JFET would only add more capacitance to the high freq problem... So, I'm going to go ahead and try 2 parallel devices but, driven with the low buffer source impeadance and isolated from the feedback network resistors... Like ZV4
Maybe nobody will notice.
New Name: ZV7-TJBC
Well, I fired the new version up. Right off the bat I knew I had a winner. Problem is, it requires considerably more voltage to get into operating range--I'm up to 20V vs. 12 with the jfet only. The good news, the 2" x 4" x 1.5" (we're talking tiny) heatsinks aren't even warm; tepid is more like it. I've only had it on for 15 min. or so, but expected more heat.
This probably means I'll have to drop the resistance beneath the CCS. Right now it's 1 ohm on a 20 watt resistor. The resistor is barely warm.
There's 1.5 volts at the source pin of the jfet. That's down from the 3 volts I had without the cascode. Now, the jfet heatsink is dead cold. The drain pin is somewhere around 4.5 volts. I know that Nelson keeps the voltage low in order to keep the jfet in that "sweet triode" range. This must mean that I'll need more current, as mentioned in the previous paragraph.
If I turn up the CCS bias pot, the sound goes bad in a heartbeat; there's probably not enough current being supplied.
The very best part is that the "golden ears" in the family said there's no contest between the sound of the new configuration and the ZV7-TB (buffered) driving the other channel. BTW, that buffer really does help this amp pull out the delicate details in the HF. I'm so sold on this baby!
This has to be, hands down, the best experiment I've ever played around with. (OK, the AX was a blast, too.) Oh, and then again, there was Mary when we were kids...
John
Well, I fired the new version up. Right off the bat I knew I had a winner. Problem is, it requires considerably more voltage to get into operating range--I'm up to 20V vs. 12 with the jfet only. The good news, the 2" x 4" x 1.5" (we're talking tiny) heatsinks aren't even warm; tepid is more like it. I've only had it on for 15 min. or so, but expected more heat.
This probably means I'll have to drop the resistance beneath the CCS. Right now it's 1 ohm on a 20 watt resistor. The resistor is barely warm.
There's 1.5 volts at the source pin of the jfet. That's down from the 3 volts I had without the cascode. Now, the jfet heatsink is dead cold. The drain pin is somewhere around 4.5 volts. I know that Nelson keeps the voltage low in order to keep the jfet in that "sweet triode" range. This must mean that I'll need more current, as mentioned in the previous paragraph.
If I turn up the CCS bias pot, the sound goes bad in a heartbeat; there's probably not enough current being supplied.
The very best part is that the "golden ears" in the family said there's no contest between the sound of the new configuration and the ZV7-TB (buffered) driving the other channel. BTW, that buffer really does help this amp pull out the delicate details in the HF. I'm so sold on this baby!
This has to be, hands down, the best experiment I've ever played around with. (OK, the AX was a blast, too.) Oh, and then again, there was Mary when we were kids...
John
Re: New Name: ZV7-TJBC
I had mine without buffer up running at more than double the voltage with good results. The harder you push it, the better it sounds....till it goes bang
Magura
carpenter said:Well, I fired the new version up. Right off the bat I knew I had a winner. Problem is, it requires considerably more voltage to get into operating range--I'm up to 20V vs. 12 with the jfet only. The good news, the 2" x 4" x 1.5" (we're talking tiny) heatsinks aren't even warm; tepid is more like it. I've only had it on for 15 min. or so, but expected more heat.
This probably means I'll have to drop the resistance beneath the CCS. Right now it's 1 ohm on a 20 watt resistor. The resistor is barely warm.
John
I had mine without buffer up running at more than double the voltage with good results. The harder you push it, the better it sounds....till it goes bang
Magura
Ok, I dropped the resistance beneath the CCS from 1 ohm to 0.5 ohm and received far better results.
Still, even after an hour of playtime, the heatsinks are only registering 32 degrees C.
All other voltages are the same. I'm keeping the supply at 20 volts for this "Little Horn Amp".
Almost forgot, I also added a feedback loop back to the buffer inputs--sound is smoother than a halo.
Comparing the newer amp to the original: both are wonderful. The original has an edge that seems to attack the music. Great for rock and roll, or music that screams. The jfet version is much sweeter, laid-back, perfect for classical and easy listening. What's amazing is that the jfet version does everything my original version does, but without the heat and on 1/3 less voltage. I'm certain that with a bit of persistence, I can get that voltage up and turn those heatsink red, but for my horns, it won't be necessary.
So, in the end, I'm keeping both.
Still, even after an hour of playtime, the heatsinks are only registering 32 degrees C.
All other voltages are the same. I'm keeping the supply at 20 volts for this "Little Horn Amp".
Almost forgot, I also added a feedback loop back to the buffer inputs--sound is smoother than a halo.
Comparing the newer amp to the original: both are wonderful. The original has an edge that seems to attack the music. Great for rock and roll, or music that screams. The jfet version is much sweeter, laid-back, perfect for classical and easy listening. What's amazing is that the jfet version does everything my original version does, but without the heat and on 1/3 less voltage. I'm certain that with a bit of persistence, I can get that voltage up and turn those heatsink red, but for my horns, it won't be necessary.
So, in the end, I'm keeping both.
I bypassed the 0.5 resistors beneath the jfet source pins and finally got the heatsinks up to 47 degrees C. The sound is not quite as silky, but the highs are more apparent. The volume increased at least 3db, maybe more.
It's obvious that I'm flowing more current. Seems to me, the voltages surrounding the jfet dropped a tad, also.
Magura, your suggestion worked great in this application. Thanks. If I have two or three more parallel transistor combinations, I will have to add a little bit of resistance to prevent current hogging. I was reading ZV3 this morning and noticed that Nelson uses 0.1 ohm resistors in parallel combinations.
Another thing; I had to reconfigure the input resistance in order for my electronic x-over to work with this amp. I dropped the 47.5k down to 1k and lowered the 221k down to 22.1k. At this moment, the 22.1k resistors are in front of the 1k resistors. I'll move them behind the 1k resistors later. The odd thing is that my little single ended input/ balanced output circuit drove this amp like no body's business. Go figure.
Cool.
It's obvious that I'm flowing more current. Seems to me, the voltages surrounding the jfet dropped a tad, also.
Magura, your suggestion worked great in this application. Thanks. If I have two or three more parallel transistor combinations, I will have to add a little bit of resistance to prevent current hogging. I was reading ZV3 this morning and noticed that Nelson uses 0.1 ohm resistors in parallel combinations.
Another thing; I had to reconfigure the input resistance in order for my electronic x-over to work with this amp. I dropped the 47.5k down to 1k and lowered the 221k down to 22.1k. At this moment, the 22.1k resistors are in front of the 1k resistors. I'll move them behind the 1k resistors later. The odd thing is that my little single ended input/ balanced output circuit drove this amp like no body's business. Go figure.
Cool.
Well, well...
I ended up calculating the amperage from the 0.5 ohm CCS source resistor and concluded that it offered only half the amperage I needed for a pair of LU1014D jfets. Like any red-blooded American, I paralleled another 0.5 ohm resistor to the original.
WOW
Of course, I now have a blow-torch sitting on my workbench. Oh well.
I had this thought:
Since I now had another fire-breathing dragon for a friend, what's to keep me from switching out the IRF610? After all, they were now passing nearly full capacity in wattage and are part of the matched set for my Russ White bosoz kit. I'd hate to kill the pair and have to purchase new parts.
IRF044N at your service. Yes, it's a wonderful transistor--more detailed sounding than the IRF610, too. Way more gain, as an added bonus.
I shut the amp down, made the swap, and after an initial adjustment and two songs to test the sonic qualities I shut her down again. Did I mention WOW ?
And yes, it runs hot. The "Little Horn Amp" heatsinks were at 60 degrees C. when I hit the "off" switch. Those were a wonderful ten minutes of listening, though.
John
I ended up calculating the amperage from the 0.5 ohm CCS source resistor and concluded that it offered only half the amperage I needed for a pair of LU1014D jfets. Like any red-blooded American, I paralleled another 0.5 ohm resistor to the original.
WOW
Of course, I now have a blow-torch sitting on my workbench.
Oh well.I had this thought:
Since I now had another fire-breathing dragon for a friend, what's to keep me from switching out the IRF610? After all, they were now passing nearly full capacity in wattage and are part of the matched set for my Russ White bosoz kit. I'd hate to kill the pair and have to purchase new parts.
IRF044N at your service. Yes, it's a wonderful transistor--more detailed sounding than the IRF610, too. Way more gain, as an added bonus.
I shut the amp down, made the swap, and after an initial adjustment and two songs to test the sonic qualities I shut her down again. Did I mention WOW
?And yes, it runs hot. The "Little Horn Amp" heatsinks were at 60 degrees C. when I hit the "off" switch. Those were a wonderful ten minutes of listening, though.
John
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.