There's no reason not to use leadfree solder, the problem stems from the fact hobbyists using $20 soldering irons from ebay which are not up to the task of providing higher temperatures that leadfree requires.
The other issue is using components down to 0805/2012 requires smaller diameter solder typically <0.3mm and very fine soldering iron tips, larger diameter solder will just cause blobbing.
Encountering SMD for the first time or people with limited experience will find the soldering experience frustrating without the right Soldering Iron/Tips and correct size/type solders. Hobbyists will buy the cheapest soldering irons and solder.
Components smaller than 1206/3216 will be a challenge for the casual hobbyist.
The other issue is using components down to 0805/2012 requires smaller diameter solder typically <0.3mm and very fine soldering iron tips, larger diameter solder will just cause blobbing.
Encountering SMD for the first time or people with limited experience will find the soldering experience frustrating without the right Soldering Iron/Tips and correct size/type solders. Hobbyists will buy the cheapest soldering irons and solder.
Components smaller than 1206/3216 will be a challenge for the casual hobbyist.
I use the smallest chisel tip you can get.
You could take your regular gauge solder and draw/pull it to a thinner diameter rather than buy a spool of the smaller gauge in order to save some $.
You could take your regular gauge solder and draw/pull it to a thinner diameter rather than buy a spool of the smaller gauge in order to save some $.
I think at least for SMT beginners and less skilled DIYers, leaded solder should be the first choice for some of its worthy qualities,
1. its eutectic behavior far more forgiving than lead-free to hand stability, soldering technique, and tools. Easy to work, easy to rework.
2. superior fluidity and "wetting" property, less prone to dry and cold joints
3. superior mechanical property, resilience to thermal cycling, stress cycling and shocks, long term reliability
4. does not require high Tg PCB, saves cost.
It would be a shame if the 'ecosystem' doesn't come to fruition. I was looking at this being a stepping stone into the world of SMD. Fingers crossed it gets resolved.
Designs are done and waiting to be tested. Files are in the first post. What needs to be resolved?
I sent Pete a bunch of design info on protection systems. He may be designing something there, in which case he may be a while.
I sent Pete a bunch of design info on protection systems. He may be designing something there, in which case he may be a while.
Well I for one was confused as to whether the boards were vetted and good to go. My knowledge is non existent in the world of SMD.
Also, there was mention of bluetooth, power meters etc. in the first post. That was what I was referring to. I'm here to learn and enjoy. Richard
Also, there was mention of bluetooth, power meters etc. in the first post. That was what I was referring to. I'm here to learn and enjoy. Richard
Pete's designs are always a group effort. He designs it, others build and test. A couple people have ordered boards already so someone should have them running soon.
Bluetooth, power meters, ect are all part of the protection system. I have a bunch of proven designs already for this. I plan to do an upgraded front panel display if time ever permits, but Pete may already be working on this.
Bluetooth, power meters, ect are all part of the protection system. I have a bunch of proven designs already for this. I plan to do an upgraded front panel display if time ever permits, but Pete may already be working on this.
I'm back ... I took a vacation from the drudgery of BOM's .. he he !
Everything I have posted should really work good.
On at least the Hellraiser and Spooky symmetrical designs , getting the base CCS current in the ballpark is the only
slight issue I envision. with all the BOM's , I have included 3 pairs of CCS emitter resistors (extra) to facilitate "getting in the ballpark".
PS - I'm not always near this PC ... the world has many demands.
OS
Everything I have posted should really work good.
On at least the Hellraiser and Spooky symmetrical designs , getting the base CCS current in the ballpark is the only
slight issue I envision. with all the BOM's , I have included 3 pairs of CCS emitter resistors (extra) to facilitate "getting in the ballpark".
PS - I'm not always near this PC ... the world has many demands.
OS
Hellraiser BOM is in the package. Chose Toshiba for the (dual) diamond and the low voltage singles. TBC847/857 are same as the dual
HN1B01 Toshiba , same process -different package. Having Q4-5 , Q8-9 the same is important on this design. I'm only using the 1845/992's for
cascodes and regulators (they need 120V devices) .... most else is Toshiba.
The current feedback resistors are good for most at 1W. I did see at >70V rail usage the dissipation is over a watt. No thin-films at 2W !!
Only 1 thick-film at 2W that fits the 2010 SMD package. I would go for the 2W if I was making a 300W amp.
OS
HN1B01 Toshiba , same process -different package. Having Q4-5 , Q8-9 the same is important on this design. I'm only using the 1845/992's for
cascodes and regulators (they need 120V devices) .... most else is Toshiba.
The current feedback resistors are good for most at 1W. I did see at >70V rail usage the dissipation is over a watt. No thin-films at 2W !!
Only 1 thick-film at 2W that fits the 2010 SMD package. I would go for the 2W if I was making a 300W amp.
OS
On to the "Mini" Arcwelder OPS (BOM). I noticed on the Wolverine they added base-stoppers to the driver stage.
A base-stopper is useful at establishing output current sharing and nullifying any tendency of negative resistance as a function
of frequency. Since the pre-drivers and drivers are (separately) decoupled and running class A , this is not needed.
Mini is good to go - BOM is next. "Mini" should be safe with just fuses and basic DC/current protection.
I noticed the larger 4 pair Wolverine still uses TO-220 drivers and is the same as the smaller version. 4 pairs is right "on the edge" of
what the MJE15032/33 pair can supply.
Since we already have 4 pair Wolverines to go to for notable power , I might make the big "arcwelder" .... 6-7 pairs - WHY NOT ? A proven 500W
output stage. I would be good to keep the staggered genders and just make a 350mm - 380mm board to use on a 400mm 3-5U heatsink.
For a board like this , it would be nice to add the MOSFET rail devices for safety and have complete integration of the protection.
There will be a full 80-100mm extra room on a huge OPS like this , Enough to integrate both current and DC on board.
Hopefully , JW can offer some hints on devices and layout. Gonna need 20-30A low Rds on high voltage devices for this.
OS
A base-stopper is useful at establishing output current sharing and nullifying any tendency of negative resistance as a function
of frequency. Since the pre-drivers and drivers are (separately) decoupled and running class A , this is not needed.
Mini is good to go - BOM is next. "Mini" should be safe with just fuses and basic DC/current protection.
I noticed the larger 4 pair Wolverine still uses TO-220 drivers and is the same as the smaller version. 4 pairs is right "on the edge" of
what the MJE15032/33 pair can supply.
Since we already have 4 pair Wolverines to go to for notable power , I might make the big "arcwelder" .... 6-7 pairs - WHY NOT ? A proven 500W
output stage. I would be good to keep the staggered genders and just make a 350mm - 380mm board to use on a 400mm 3-5U heatsink.
For a board like this , it would be nice to add the MOSFET rail devices for safety and have complete integration of the protection.
There will be a full 80-100mm extra room on a huge OPS like this , Enough to integrate both current and DC on board.
Hopefully , JW can offer some hints on devices and layout. Gonna need 20-30A low Rds on high voltage devices for this.
OS
I add the rail mosfets to the power supplies before the fuses in my designs. I use the same mosfets as the speaker relays use. IPB025N-10N3 G or whatever is available with similar specs. Go for 150 amp+ for lower RDSon. Should likely be able to withstand double the rail voltage, but the rails are disconnected within a couple milliseconds of any sign of trouble so there likely wouldn't be much chance of opposite rail voltage getting back to them.
So just add it to my power supply board ? Yes , I'll just replace those 24V supplies with the rail MOSFET's. For any comprehensive
protection scheme and/or any input conditioning , a separate trafo and supply would be superior.
I like keeping the SMD and through-hole separate , I'll use TH MOSFETS and opto's on the PS board , keep SMD digital on the protection PCB.
The reason is that some might want the glory of assembling the OPS and PS boards themselves while outsourcing prebuilt boards from us "SMD guys" ! he he....
Let's see , a (IPP180N10N3GXKSA1) is a "beefier" 43A/205A TO-220 device - cheap at $1.31 ... Why only N-channel devices ? No P/N for the 2 rails ?
ASSR-622 opto has a integrated "turn-off" circuit. "Turn-off circuit" ? Just an external gate-source resistance ?
OS
protection scheme and/or any input conditioning , a separate trafo and supply would be superior.
I like keeping the SMD and through-hole separate , I'll use TH MOSFETS and opto's on the PS board , keep SMD digital on the protection PCB.
The reason is that some might want the glory of assembling the OPS and PS boards themselves while outsourcing prebuilt boards from us "SMD guys" ! he he....
Let's see , a (IPP180N10N3GXKSA1) is a "beefier" 43A/205A TO-220 device - cheap at $1.31 ... Why only N-channel devices ? No P/N for the 2 rails ?
ASSR-622 opto has a integrated "turn-off" circuit. "Turn-off circuit" ? Just an external gate-source resistance ?
OS
The mosfets are switched with specialized opto-isolators. The optos produce enough voltage internally to switch on the mosfets and they actually produce a small momentary negative charge for "fast off" This is a floating circuit so it can be installed in either a positive or a negative rail, just reverse the drain and source to match. Each rail gets it's own opto channel.
I like to keep the digital stuff completely isolated from the audio. No shared power or grounds anywhere. This eliminates any possible ground loops being caused by the protection system and no chance of 16 MHz or whatever clock speed you're running from getting into the amp through the supplies.
I like to keep the digital stuff completely isolated from the audio. No shared power or grounds anywhere. This eliminates any possible ground loops being caused by the protection system and no chance of 16 MHz or whatever clock speed you're running from getting into the amp through the supplies.
So it is a specialized circuit to negate the gate capacitance.
They do make a TH DIP of this device - https://www.mouser.com/ProductDetail/Broadcom-Avago/ASSR-V621-002E?qs=7zQgkplsu/MyPB9VMjuAnA==
ASSR-622 + 2 TO-220 Infineon MOSFETS = a switchable PS. The serial control chip can stay on the digital (SMD) board.
As a joke , shall we make our circuit "AI enhanced" ?
The a$$hole's just integrated AI into my Opera browser after an update. I DON'T want AI on this PC !
PS - not for my benefit , they just want me to interact and train it for free = "Scamerica" !!
OS
They do make a TH DIP of this device - https://www.mouser.com/ProductDetail/Broadcom-Avago/ASSR-V621-002E?qs=7zQgkplsu/MyPB9VMjuAnA==
ASSR-622 + 2 TO-220 Infineon MOSFETS = a switchable PS. The serial control chip can stay on the digital (SMD) board.
As a joke , shall we make our circuit "AI enhanced" ?
The a$$hole's just integrated AI into my Opera browser after an update. I DON'T want AI on this PC !
PS - not for my benefit , they just want me to interact and train it for free = "Scamerica" !!
OS
Why use a bloated browser owned by Beijing Kunlun (a company with ties with the CCP?) You can find several open source secure browsers.
There are loads of optos designed to drive mosfets. Those are just the ones I ran with. One hitch with them is they make an identical one with the output pin connections reversed that can cause some confusion.
Yes a simple 2 wire to switch the supply on works. I just used the I2C control because I had a I2C bus running around the amp for temperature sensing already.
There are very cool current limiting devices to drive LEDs and optos with. Basically current limiting diodes that let you drive the optos with up to 45 VDC so it allows you to switch it with anything. NSI45015WT1G is what I've been using. Hopefully available in a through hole version, but like any cool, super easy to implement device, not likely.
https://www.digikey.ca/en/products/detail/onsemi/NSI45015WT1G/2401410
Yes a simple 2 wire to switch the supply on works. I just used the I2C control because I had a I2C bus running around the amp for temperature sensing already.
There are very cool current limiting devices to drive LEDs and optos with. Basically current limiting diodes that let you drive the optos with up to 45 VDC so it allows you to switch it with anything. NSI45015WT1G is what I've been using. Hopefully available in a through hole version, but like any cool, super easy to implement device, not likely.
https://www.digikey.ca/en/products/detail/onsemi/NSI45015WT1G/2401410
"Bloated browser" 1/10 google chrome. Then there is Vivaldi and the lightest (Brave) webkit based one. My point is that AI is getting pushed on us
whether we want it or not. Edge has it integrated. At least i can turn it off , even as I am sure it is still datamining (for training).
Looks good , is the improved shutoff time critical in a protection scenario ? I mean versus a plain gate bleeder resistance ? No issue that it costs $4.50 , as
it is optimal for the job and far cheaper than blown speakers.
OS