jp_howard said:
Cordless drill for drilling hole
and a manual "t-piece" for the tap.
A drill for tapping a thread?
will either strip or break tap.
On the M3 tap package, it recommends a 2.5mm drill bit.
Hugh
"Use a cordless drill on slow speed with lowest torque setting."
for the tap or drilling the hole?
And Jeremy
Remmber changing the Re's from 0.22 to 0.1 with change the bias, ie lower resistance, more current.
lower the bias before you change the emitter resistors.
Remmber changing the Re's from 0.22 to 0.1 with change the bias, ie lower resistance, more current.
lower the bias before you change the emitter resistors.
Yeeessss.... weeelllll...... I admit it, I've already been bitten by this once - the source of one more flaming SCULD.awpagan said:
So I hopefully won't make the same mistake again!Master Hugh told me that creating an amplifier graveyard is a necessary step in the learning process. So does that mean the amplifier massacre I've been engaged in over the last week mean I must be very educated by now? 😉
jp_howard said:
Does that mean your've popped an electrolytic cap yet?
amplifier massacre/graveyard!!! hmmmm spare parts/parts bin 😀
jp_howard said:
JP
Do you mind sharing these instructions?
I keep screwing up (also an intended pun) everytime...
Thanks!
jp_howard said:
Right on - not only the ones that blew-up along the way, but also those that "work" (and sound awful). The nice thing about it is that you now have your own surplus supply. 😀
The combination of persistence + education is key in my life - DIY audio is no exception.
Cheers!
Not at all - but they're not mine to share. It was a private email from Alfred. However, I know he's following this thread, so hopefully he'll share his wisdom here.dudaindc said:
Hey that's my joke - get your own! 😉
Is this the post about Alfred's mods
http://www.diyaudio.com/forums/showthread.php?postid=1789578#post1789578
I've done a few of these changes, plus something else.
Will post a modified schematic later of what i've done.
just for comparison and views
http://www.diyaudio.com/forums/showthread.php?postid=1789578#post1789578
I've done a few of these changes, plus something else.
Will post a modified schematic later of what i've done.
just for comparison and views
FWIW:
British Standard tap drill size for a 3.0mm coarse thread is 2.5mm, or 2.55 with less radial engagement. I guess if you use 2.7mm, the likelihood of tap breakage is much less, but so is the engagement.
I find the easiest way to ensure full, square threads this small is to fit the tap into the bench drill chuck, and the work on the table, in a vise if necessary. Don’t power on; rotate the chuck by hand while gently pulling on the lever. I usually carry on this way until the required thread depth is achieved, but once the thread is well started, you could of course release the tap from the chuck with the tap still in the material, and carry on with the tap wrench.
For aluminium, my cutting lubricant is common or garden white spirit.
Buy a plug tap to get the thread all the way down the hole, or grind away the taper on a normal taper tap (same thing but shorter).
And I don’t often use 3mm screws; I’ve still got hundreds of 6BA. 🙂
Brian.
British Standard tap drill size for a 3.0mm coarse thread is 2.5mm, or 2.55 with less radial engagement. I guess if you use 2.7mm, the likelihood of tap breakage is much less, but so is the engagement.
I find the easiest way to ensure full, square threads this small is to fit the tap into the bench drill chuck, and the work on the table, in a vise if necessary. Don’t power on; rotate the chuck by hand while gently pulling on the lever. I usually carry on this way until the required thread depth is achieved, but once the thread is well started, you could of course release the tap from the chuck with the tap still in the material, and carry on with the tap wrench.
For aluminium, my cutting lubricant is common or garden white spirit.
Buy a plug tap to get the thread all the way down the hole, or grind away the taper on a normal taper tap (same thing but shorter).
And I don’t often use 3mm screws; I’ve still got hundreds of 6BA. 🙂
Brian.
I haven't seen something like that before - there's 2 separate windings attached to different bits of the circuit, is there? What does that do? How does it work?
Does the JLH add-on go on both rails? Are there 2 separate power supplies?
Does the JLH add-on go on both rails? Are there 2 separate power supplies?
Today's update: replaced the emitter R's with 0R1's this morning (from 0R22's). I can't hear a difference, but the voltages are now less well matched and less stable. I've commented on this further in the 'ULD extreme' thread.
Since that didn't do anything much, this arvo I did the current mirror stuff - put in bc550c's with HFE>500, fixed the trannie matching in my 2nd amp using the extra trannies I bought yesterday, and adjusted the trimpot as described by Alex to get 0mV across the collectors (actually, 0.2mV, but that'll do I guess...)
I haven't done enough listening tests to comment on the result yet. Lots of listening to do tomorrow!
In case anyone's interested: I found quite an efficient way to do the transistor matching. I had 25 transistors of each type, so I put a 5x5 grid on a piece of foam - one piece per transistor type. I wrote the numbers 1-25 in the grid cells. I then opened a spreadsheet and put the numbers 1-25 down the first column. Then I put on my headset, and used voice recognition (e.g. Dragon Naturally Speaking, Vista speech recognition, or MS Office recognition) to speak the relevant measurements of each device as I tested them. After testing, I pressed the trannie into the foam in the next grid position. This saved lots of time, because I never had to move from my test equipment to write down the values - the process was just: grab trannie, test, speak, stick in foam. When I was done for one trannie type, I simply used the 'sort' command in the spreadsheet to sort by whatever I wanted to match on, and used the item numbers (1-25) in the first column to grab a pair from the foam. The foam pieces seem to be a good way to keep the remaining transistors safely; I've pasted the measurements to each one (with the removed trannies crossed out) so for the next project I'll be ready to go.
When replacing some transistors in the PCB, I must have used too much heat, because 2 of the pads became unstuck. 😡 What's the right way to fix this, after it's happened? In my case, I tried to get the best join I could, and then scraped back the track to show the copper underneath, and finally used a 'circuit writer' pen to draw over the copper and the component's lead. I'm not sure how reliable this will be in the long term. It seems to be working OK for now however...
Since that didn't do anything much, this arvo I did the current mirror stuff - put in bc550c's with HFE>500, fixed the trannie matching in my 2nd amp using the extra trannies I bought yesterday, and adjusted the trimpot as described by Alex to get 0mV across the collectors (actually, 0.2mV, but that'll do I guess...)
I haven't done enough listening tests to comment on the result yet. Lots of listening to do tomorrow!
In case anyone's interested: I found quite an efficient way to do the transistor matching. I had 25 transistors of each type, so I put a 5x5 grid on a piece of foam - one piece per transistor type. I wrote the numbers 1-25 in the grid cells. I then opened a spreadsheet and put the numbers 1-25 down the first column. Then I put on my headset, and used voice recognition (e.g. Dragon Naturally Speaking, Vista speech recognition, or MS Office recognition) to speak the relevant measurements of each device as I tested them. After testing, I pressed the trannie into the foam in the next grid position. This saved lots of time, because I never had to move from my test equipment to write down the values - the process was just: grab trannie, test, speak, stick in foam. When I was done for one trannie type, I simply used the 'sort' command in the spreadsheet to sort by whatever I wanted to match on, and used the item numbers (1-25) in the first column to grab a pair from the foam. The foam pieces seem to be a good way to keep the remaining transistors safely; I've pasted the measurements to each one (with the removed trannies crossed out) so for the next project I'll be ready to go.
When replacing some transistors in the PCB, I must have used too much heat, because 2 of the pads became unstuck. 😡 What's the right way to fix this, after it's happened? In my case, I tried to get the best join I could, and then scraped back the track to show the copper underneath, and finally used a 'circuit writer' pen to draw over the copper and the component's lead. I'm not sure how reliable this will be in the long term. It seems to be working OK for now however...
When testing changes at Hugh's, life was made a lot easier by the fact that he has a variable autotransformer (i.e. like a Variac - but of his own design). It's probably not a viable option for me - big, expensive, and hard to find.
In looking around for alternatives, the only thing I've found is the 'dim bulb tester'. This is simply a regular household bulb in series before the transformer, with a switch to bypass it as required. Different wattage globes can be inserted to change the power to the circuit.
Is this a good method to test an amp without letting the smoke out? If so - does anyone have any tips on a suitable design (or is it just as simple as I described)? If not, are there any other alternatives? (The SC article suggests removing the fuses and sticking 0R68's there instead for testing, but that gets rather tiresome, and in my experience it's not enough to keep the smoke bottled up all the time).
In looking around for alternatives, the only thing I've found is the 'dim bulb tester'. This is simply a regular household bulb in series before the transformer, with a switch to bypass it as required. Different wattage globes can be inserted to change the power to the circuit.
Is this a good method to test an amp without letting the smoke out? If so - does anyone have any tips on a suitable design (or is it just as simple as I described)? If not, are there any other alternatives? (The SC article suggests removing the fuses and sticking 0R68's there instead for testing, but that gets rather tiresome, and in my experience it's not enough to keep the smoke bottled up all the time).
jp_howard said:
Personally I dont use the light bulb method.
I wire up the amp in stages checking each stage before and after it is connected.
I check the power supply first.
I then add the driver board with the VAS output fed back into the LTP.
I then make the bias as low as possible.
Only then would I connect the output transistors.
I always use the bulb tester when starting up any and every modification to a piece of equipment.
To date I have never blown up any circuit with the bulb tester in place.
I have blown up more than a few when probing with a DMM and touching/shorting the wrong lead/s.
If you look at the RMI thread, you'll see that I have a motorboating problem with the bulb tester in place. The motorboating may be an interaction between the PTC of the filament and the current/capacitance in the RMI. But, importantly, it has not been damaged.
So far out of hundreds of bulb protected switch ons I have never damaged anything.
To date I have never blown up any circuit with the bulb tester in place.
I have blown up more than a few when probing with a DMM and touching/shorting the wrong lead/s.
If you look at the RMI thread, you'll see that I have a motorboating problem with the bulb tester in place. The motorboating may be an interaction between the PTC of the filament and the current/capacitance in the RMI. But, importantly, it has not been damaged.
So far out of hundreds of bulb protected switch ons I have never damaged anything.
Thanks Andrew - that's very interesting. Could you tell me how you've got that set up? Just one bulb in series before the transformer? What wattage? Any other details I should know about before I set this up?
What is 'motorboating'? I've googled without any success - lots of information about boats with motors, but nothing about circuits!...
jp_howard said:
The separate front end psu was used in the earlier version "ULD100W" were there wasn't the ccs circuit for power supply ripple rejection.
The transformer had 2*35Vac secondary windings and an extra 2*15Vac 500mA windings.
the SC100Watt circuit is from there
http://www.diyaudio.com/forums/showthread.php?s=&threadid=103157&highlight=
The 35vac and 15vac and run in series to get 50vac then reg'd back to +/-55Vdc for the front end.
The outputs are not reg'd, just use the 35vac >bridge >caps to get the 52Vdc
Yes i modded the JLH for this too.
Jaycar did have the kit, but no longer listed
Altronics looks like they still do, but the pic is wrong🙄
http://www.altronics.com.au/index.asp?area=item&id=K5152
As the pad lift, could use a short link wire back track
Attachments
jp_howard said:
Good description be ostripper
http://www.diyaudio.com/forums/showthread.php?s=&postid=1769670&highlight=#post1769670
- Status
- Not open for further replies.