Balanced - Cascoded F5

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I was fairly far along with my Cascoded F5, and I was in the midst of my Balanced F5 when I realized a trivial change could transform one of my projects to a Balanced Cascoded F5.

I completed the transformation on one channel tonight. The transformed board is not as pretty as I'd hoped, but I'll make a set of custom boards if I like the outcome. Believe it or not, I literally ran out of wire during the build, so testing will have to wait until tomorrow.

On another note, I was also the recipient of a kind gift - mica insulators (thanks to TB). Now my heat sink measurements should be better.

In any case, I hope the outcome is as good as the excitement I have in building the project.
 
What voltages have you decided to go with?

I wouldn't have thought that this was a particularly difficult question to answer. Particularly in light of the fact that you were complaining about other people not answering your questions, I find this rather strange behaviour.

Even a "I don't know" would have sufficed.

I have lost all sympathy.
 
Last edited:
Melon,

I know you're eager to see where this goes, so am I. I think you need to realize I have kids, my weekends are not my own. I haven't checked this thread until this minute because I haven't had anything new to add.

On one of your points: the difference between me and the folks I am not impressed with is - I will answer any question I can to the fullest extent possible, the others met out information like rationed food in wartime.

Finally, I haven't decided on any voltage yet. I've converted the board and have one channel ready for testing. I have two toroids at the moment - 18V and 26V, I'll try them both.

I hope to get time tonight to press on, but weekends are for the kids. Patience, I think I've gone from design to built channel and ready to test in something like 10 days. Give me a day or two to find time to fire up the beast!

I wouldn't have thought that this was a particularly difficult question to answer. Particularly in light of the fact that you were complaining about other people not answering your questions, I find this rather strange behavior.

Even a "I don't know" would have sufficed.

I have lost all sympathy.
 
Melon,

I know you're eager to see where this goes, so am I. I think you need to realize I have kids, my weekends are not my own.

I hope to get time tonight to press on, but weekends are for the kids. Patience, I think I've gone from design to built channel and ready to test in something like 10 days. Give me a day or two to find time to fire up the beast!

Having children myself I understand your predicament completely.
No problem
 
It's alive!

I finally had a chance to try the beast. It lives!

I used my stock F5 toroid 18V 400VA for 22.4V rails (measured).

I had an interesting time biasing. I'd made the mistake (the last time I do such) of buying trimmers from Tayda. They sent 100K trimmers instead of the 5K ordered. Now I didn't know if the spaghetti mix of cross connections was correct, and every time I turned a trimmer, Pow! - amps would flow like Niagara Falls. I kept thinking I'd made a connection mistake.

In the end, I checked the trimmers - they were 100K which accounted for the wild changes with each turn. Luckily I had some 2Ks around.

The next adventure was again my fault. Another forum member had recommended grounding the "X" for the initial bias. I tried this and couldn't get stability between on-off cycles. I un-grounded the "X" and the circuit was rock solid.

I biased to 2.6A for each side of the circuit (like the stock F5) making the Balanced side 5.2A. This is too much for my Conrads - they settled in at 30C above room temperature.

I cut the bias to 4A for the channel and got 26C above room temperature. Still a little high.

Tomorrow brings sound. I'm hopeful for a nice outcome. In the end I have to make some choices about voltage bias and heat-sink size. I think I'd have gone with bigger sinks if I knew I was going a Balanced Cascoded F5.

Ironically, this is the circuit I proposed a few times in the forum when I first started out (except for the cascoding).
 
Member
Joined 2006
Paid Member
I would think if your gonna cascode, then might as well spike up the rail voltage.
I've got a BA2 cascoded in the works with 12 outputs per channel. Still waiting for some parts to come.

I am watching your threads here to wait for some sonic results of the cascoding and higher voltage input.

One can't be too eager. I've got a kid too, who is sick this weekend, so science project's stop. Good luck.

Btw, I like the cheaper Vishay T64 pots, as the resistance is printed clearly on it.

Good luck.
 
I'm happy to be wrong.

Here's what I have -

Each channel of the circuit for the Balanced - Cascoded F5 has four outputs, two IRFP240s and two IRFP9240s. I measured the voltage across the 0R47 resistor at each output device. I measured 0.470V at each of four outputs to yield the following:

Amps = 0.470/.47 = 1 per output. I have four outputs, each with 0.470V across the 0R47 resistor, 4A right?

In the stock F5, you bias to 0.6V on each output and get 2.6A total. I assumed each output is additive.

As I've said before, I'm relatively new to this hobby, perhaps I don't fully understand the implications of the circuit. That's why I like to experiment. I learn.

this cannot be right.
The bias for a standard F5 is 1.3A, not 2.6A

I suspect you have not read that other thread yet!
 
I assumed each output is additive.

Each output PAIR is additive but NOT each output.

This is good news.
If you have 22V rails you can safely increase the bias to around 100W dissipation.
P=VI

I=100/(2x22) = 2.27A.
You have two output pairs. That is 1.135A through each source resistor.

Don't go higher than 130W. That is don't go higher than 2.95A total bias.
So if you are happy with the heat, you can safely take it to 2.6A (ie 1.3A per output pair).
 
Last edited:
I finished one channel, then converted the board to this project. If (or when) I un-"X" this board, it will be two cascoded channels for an F5. I'll get to all of this the first chance I have.

The plan is to try the Balanced Cascoded F5 channel sonically tonight (hopefully), then un-"X" it and try the Cascoded F5, then try the Stock F5 (different build). In the background, I have a Balanced (not cascoded) channel set up for the Toshiba outputs that I will also try sonically - but one must have the time for all these experiments. I suspect by this time next week, I'll have tried them all at different settings.

Depending on the first listening session, I'll likely order a few more toroids - I'm considering 12V for the Balanced and ~28-32V for the Cascoded. I haven't thought too much about the ideal for the Balanced Cascoded, but I'm sure the answer will come.

BTW, if anyone is in the NYC area and wants to come by, PM me. If you have different toroids, bring them along!
NYCOne,
Did you ever finish your Cascoded F5 ? if so what were your final impressions ?


regards,
 
I was trying to get the most Class A power for the given outputs. Lower voltage yields more "space" for the current when trying to limit dissipation per device to 30W.

That said, I'm not married to 12V - I could see 15V. That's what I'd planned initially.

I assume that is 12V AC not DC. Why so low?
That is around 15V DC. Probably lower once you add the crc supply
I think a torroid with 15V secondaries would be a better match.
 
All Smiles...

I hooked up my solo channel of my new Balanced Cascoded F5. The bias settings from last night were dead on tonight - a great start.

I've built a working Stock F5, and played around with various permutations of the F5, but I NEVER get tired of that first song though a new amp - pure unadulterated bliss. I chose Jet "Are You Gonna Be My Girl", followed up by a little Professor Longhair - "Tipitina"

Needless to say, the amp is wonderful. It seems similar to the Stock F5 in many ways: very detailed, excellent highs. I say this but must tell you, this is not the way to compare amps - I have only one channel, hooked to a makeshift FF85K speaker (in a Logitech enclosure). To say the least, I need to finish my OBs (all parts here, just need time).

So far, what I can say is the Balanced - Cascoded F5 is viable and worthwhile.

Tests to do when time permits:

  1. Stock F5 vs. Balanced Cascoded F5 (both with IRFP outputs, and already built)
  2. Stock F5 vs. Balanced Cascoded F5 vs. "Stock" Balanced F5 with Toshiba outputs (just need to finish the "Stock" Balanced Toshiba - one night or so)
  3. Balanced Cascoded F5 vs. "Stock" Balanced F5 (both with Toshiba outputs - need to convert Balanced Cascoded F5 to Toshibas and finish Balanced F5 - two or three nights)
  4. Cascoded F5 vs. "Stock" Balanced F5 (have all the bits, just need to build - a week or so)
  5. Balanced Cascoded F5 vs. Cascoded F5 (this is a long way off - lots of shared bits at this point)
I'm having a lot of fun. I am not well set up for technical analysis, I'll leave that to folks with the right equipment. I just like to build things and listen to music - pretty simple really.

I'd like to take a moment to point out none of this would be possible without Nelson Pass. I would like to thank him for his kindness and generosity - they are both valued deeply in NYC!
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.