resistor hunt
Jens: Thanks for the link, but there is too much K in danish for me to understand ;-). I agree that the footprint you've used is standard, but
I think that it is standard for 1/4W resistor. We normally use slightly longer footprint(11mm body instead of 7.5mm) for 1/2W, 'cause there is not a lot of manufacturer selling small footprint 1/2W namely BC comp and Phoenix. These are specialty resistors and not cheap.
Well, I guess I will end up using Xicon resistor with longer lead unless
I find a replacement solution...
Jens: Thanks for the link, but there is too much K in danish for me to understand ;-). I agree that the footprint you've used is standard, but
I think that it is standard for 1/4W resistor. We normally use slightly longer footprint(11mm body instead of 7.5mm) for 1/2W, 'cause there is not a lot of manufacturer selling small footprint 1/2W namely BC comp and Phoenix. These are specialty resistors and not cheap.
Well, I guess I will end up using Xicon resistor with longer lead unless
I find a replacement solution...
The 0.6W is the metal film resistor type power rating.
This is the same size as a 0.25W carbon mass resistor.
\Jens
This is the same size as a 0.25W carbon mass resistor.
\Jens
My personal preference is for KOA Spear over Xicon (both from Mouser), but I have no real reason for it.
Re: resistor hunt
Jean, read my post on T1 and T2 test temperatures at the Krell thread and check Dale RN60 at Mouser.
(8.75 mm body i believe)
jean.ricard said:
Jean, read my post on T1 and T2 test temperatures at the Krell thread and check Dale RN60 at Mouser.
(8.75 mm body i believe)
1/4w resistors are ok for any spot on the board that doesn't require 1/2 w or more. See Leach's web site for this.
If you can fit higher wattage ones in a spot, that's better. Theorectically less noise. Make sure that any resistor you get is dimensioned so the leads will fit in the holes.
Jens, could you give us the hole diameter specification for the resistors please?
If you can fit higher wattage ones in a spot, that's better. Theorectically less noise. Make sure that any resistor you get is dimensioned so the leads will fit in the holes.
Jens, could you give us the hole diameter specification for the resistors please?
OK I’ve got a few questions.
Since no one answered my question about the resistors listed as “TO BE DEFINED” I have been comparing the original schematic to Jen’s. This is what I found.
Jen’s R26,R34, (R30, R31 for Leach) = 3.9K 1/2W
Jen’s R28,R30, (R28, R29 for Leach) = 270R
Jen’s R41-R43, R60-R62, (R37-R40 for Leach) = 680R
Is there a reason why these values won’t work? Hopefully some of you guys that are so good with the math can give me something to go on. I’d like to get these ordered along with the other stuff, to save on shipping charges.
And then there’s Jen’s R68, R69, I can’t find these on the Leach circuit. I don’t see any resistors tying the bases of Q10 & Q11 to the rails on the Leach schematic. Am I missing something?
Thanks, Terry
Since no one answered my question about the resistors listed as “TO BE DEFINED” I have been comparing the original schematic to Jen’s. This is what I found.
Jen’s R26,R34, (R30, R31 for Leach) = 3.9K 1/2W
Jen’s R28,R30, (R28, R29 for Leach) = 270R
Jen’s R41-R43, R60-R62, (R37-R40 for Leach) = 680R
Is there a reason why these values won’t work? Hopefully some of you guys that are so good with the math can give me something to go on. I’d like to get these ordered along with the other stuff, to save on shipping charges.
And then there’s Jen’s R68, R69, I can’t find these on the Leach circuit. I don’t see any resistors tying the bases of Q10 & Q11 to the rails on the Leach schematic. Am I missing something?
Thanks, Terry
Has anybody seen a price around $4 (Unit Cost) for the MJL4302 / MJL4281 transistors. ?
I was given this price which I feel is good- but the lead time is 8 Weeks
I was given this price which I feel is good- but the lead time is 8 Weeks
Byrd said:Has anybody seen a price around $4 (Unit Cost) for the MJL4302 / MJL4281 transistors. ?
I was given this price which I feel is good- but the lead time is 8 Weeks
Digikey has them in stock, but the price doesn't come down to $4 until you order 500 of each.
Arrow has them in stock for $5 in small quantity.
Resistor holes are:
Small resistors and non inductive emitter resistors = 0.8mm.
Rest of the resistors are 1.0mm.
Regarding the TBD resistors in the protection they will depend on what rail voltage you use and the type of output transistors. Also it is a matter of how close you want to allow the transistors to get to the SOA curve.... and what SOA curve you want to use.......
Please be patient, I believe I have done my share for this project, but will return with some sort of pdf file to help you chose your resistor values. This will be done when I have the time and energy to sit down and do it. I have started something in mathcad, but it is no condition to precent to anybody yet.
You can run the amp without the protection circuit, just be carefull that you do not short the output terminals.
\Jens
Small resistors and non inductive emitter resistors = 0.8mm.
Rest of the resistors are 1.0mm.
Regarding the TBD resistors in the protection they will depend on what rail voltage you use and the type of output transistors. Also it is a matter of how close you want to allow the transistors to get to the SOA curve.... and what SOA curve you want to use.......
Please be patient, I believe I have done my share for this project, but will return with some sort of pdf file to help you chose your resistor values. This will be done when I have the time and energy to sit down and do it. I have started something in mathcad, but it is no condition to precent to anybody yet.
You can run the amp without the protection circuit, just be carefull that you do not short the output terminals.
\Jens
JensRasmussen said:
Hi Jen's
I know you have done more than your share for this project. That is why I addressed all of the folks and not just you.
Patience is a virtue
Have them if you can
Seldom found in women
Never found in men
😀
It is very tough being patient when I have four beautiful boards just sitting there on my bench.
Since I'm a total newb, may I assume you are refering to R68 and R69 when you say protection circuit or does it include all of the TBD resistors? Which ones can be left off for now?
I notice that R68, R69, C35 & C42 are not present in your 10 transistor version. Are they necesarry in this version?
For the record, I am planning on using MJL4302/MJL4281 outputs and 64V rails.
Thanks, Terry
JensRasmussen said:
22 gauge wire is supposed to be .81mm.
Holco H4 half watt resistors are specified to have .032 in diameter leads, the same as 22 gauge wire, and they will not fit the holes for the small resistors!!
Their bodies are too long, anyway, but be careful about getting into this diameter range of lead. Drilling the holes larger would be a problem with plated through holes. If you do for any reason, make sure the lead is soldered on both sides of the board.
A silly .01mm seems to make a difference, unless the Holcos I have are at the large end of their tolerance!!! Is this due to plating making the hole smaller?
The small resistors I use are 0.6 mm.
The hole size depends on plating, but usually you specify the final hole size, and the manufacturer adds whatever is necessary to arrive at the specified hole diameter. I don’t know the manufacturer in the USA, but it is common practice here in Denmark to do it this way.
The problem might have to do with conversion from mm to inches at the PCB factory. I specified the hole size in mm for the manufacturer.
In the BOM there are pics of the resistors to try to enable you to get the right parts. Should you have resistors that will fit on the PCB but not in the holes, you can always enlarge the holes and solder the resistors on both sides on the PCB.
Do not do this for parts (mostly caps) that do not have the solder pads visible on both sides when mounted!!!!
I have yet to receive my boards, so I don’t know what the final result looks like yet.
\Jens
The hole size depends on plating, but usually you specify the final hole size, and the manufacturer adds whatever is necessary to arrive at the specified hole diameter. I don’t know the manufacturer in the USA, but it is common practice here in Denmark to do it this way.
The problem might have to do with conversion from mm to inches at the PCB factory. I specified the hole size in mm for the manufacturer.
In the BOM there are pics of the resistors to try to enable you to get the right parts. Should you have resistors that will fit on the PCB but not in the holes, you can always enlarge the holes and solder the resistors on both sides on the PCB.
Do not do this for parts (mostly caps) that do not have the solder pads visible on both sides when mounted!!!!
I have yet to receive my boards, so I don’t know what the final result looks like yet.
\Jens
Hi Bowdown,
the gain of the power amp can be altered by changing the ratio of (R12 + R13) / R16 and R8 has to equal (R12 + R13).
It is easier to just change R16 since this does not affect the two feedback loops that have interlinked time constants (R8/C5 & R13/C9).
When you change R16 you also need to change C12 to reset the bass roll off frequency. Jens has set this RC time constant to 470mS but any value between 100mS and 470mS will do, however I recommend you aim for 150mS.
The gain for different R16 (lower leg feedback ) values is as follows
R16=1k0 gain = 23times. R16=750r gain = 30.3times. R16=510r gain = 44.1times.
Since Leach went to great lengths explaining that the input stage becomes nonlinear above 1000mV peak (about 700mVrms) then a sensitivity of this value will avoid the distortion that he has tried to design out of this project. This indicates to me an R16 value of 510r for a 120Watt into 8r amp.
the gain of the power amp can be altered by changing the ratio of (R12 + R13) / R16 and R8 has to equal (R12 + R13).
It is easier to just change R16 since this does not affect the two feedback loops that have interlinked time constants (R8/C5 & R13/C9).
When you change R16 you also need to change C12 to reset the bass roll off frequency. Jens has set this RC time constant to 470mS but any value between 100mS and 470mS will do, however I recommend you aim for 150mS.
The gain for different R16 (lower leg feedback ) values is as follows
R16=1k0 gain = 23times. R16=750r gain = 30.3times. R16=510r gain = 44.1times.
Since Leach went to great lengths explaining that the input stage becomes nonlinear above 1000mV peak (about 700mVrms) then a sensitivity of this value will avoid the distortion that he has tried to design out of this project. This indicates to me an R16 value of 510r for a 120Watt into 8r amp.
Hi Still4given,
the protection resistors R68 & R69 are needed to allow a dual slope protection locus.
This dual slope is needed to try to get the best out of the output transistors without the protection needlessly cutting in when protection is not actually required. Activation of the protection produces a terrible noise and many have comdemned otherwise good amps for poor sound quality because the protection was badly designed. Give Jens and others time to come up with a scheme to suit a variety of voltages / output stage transistors / temperatures(bias, heatsinks & ambient temp) / speaker load etc.
The extra capacitors (C17 & C18) reduce short voltage peaks before they get to the bases of the protection transistors. This allows short term power overloads to get through to the output without the protection cutting in. Again this would sound bad and very short overloads that would be outside the DC SOAR are still inside the 10mS or 100mS SOAR quoted on the manufacturers data sheets.
Be patient and just leave the two transistors and the resistors you have listed out of the first assembly.
the protection resistors R68 & R69 are needed to allow a dual slope protection locus.
This dual slope is needed to try to get the best out of the output transistors without the protection needlessly cutting in when protection is not actually required. Activation of the protection produces a terrible noise and many have comdemned otherwise good amps for poor sound quality because the protection was badly designed. Give Jens and others time to come up with a scheme to suit a variety of voltages / output stage transistors / temperatures(bias, heatsinks & ambient temp) / speaker load etc.
The extra capacitors (C17 & C18) reduce short voltage peaks before they get to the bases of the protection transistors. This allows short term power overloads to get through to the output without the protection cutting in. Again this would sound bad and very short overloads that would be outside the DC SOAR are still inside the 10mS or 100mS SOAR quoted on the manufacturers data sheets.
Be patient and just leave the two transistors and the resistors you have listed out of the first assembly.
AndrewT said:
There's that "Be patient" thing again. I thought I explained that.
😀 I guess I'll just wait. I'll build something else in the mean time. I have spent soooooo much money on shipping because I forgot to order something or ordered something wrong. Sometimes the shippng is as much as the parts I ordered. I just hate that.😡
I'll just keep an eye on this thread and hope to see when this is all worked out. I'm not going to order any parts until I can order all of them. I think I've already paid for one of those big brown vans since I started this hobby.
Thanks for explaining things.
Blessings, Terry
Terry,
You can use the amp without the protection circuit if you are careful.
Then piggy back the protection circuit parts on the your next project's parts order.
The Leach specified values ought to give reasonable protection at your rail voltages if you don't want to risk it. Overlay the MJ15003 SOA curve on the MJL4302 curve as a sanity check. It looks like the MJ15003 single pulse curve falls inside the 1 second curve of the MJL4302.
The double slope protection circuit Jens has built in allows use more of the SOA that you are paying so dearly for with the MJL4281/4302. The MJL3281/1302 has roughly the same SOA as the MJ15003/4 up to its 15 amp limit and costs only $2.07 at Arrow. The 4281/4302 is a more rugged part if you plan to use it on stage, but in a home/studio application the 3281/1302 should be suitable.
I wish that I could help design the protection circuit. Thanks again for all the work you've shared with us Jens.
BTW, I have tried various resistors (not Holco) and have had no problems with the leads not fitting. A snug fit in spots, but a fit.
You can use the amp without the protection circuit if you are careful.
Then piggy back the protection circuit parts on the your next project's parts order.
The Leach specified values ought to give reasonable protection at your rail voltages if you don't want to risk it. Overlay the MJ15003 SOA curve on the MJL4302 curve as a sanity check. It looks like the MJ15003 single pulse curve falls inside the 1 second curve of the MJL4302.
The double slope protection circuit Jens has built in allows use more of the SOA that you are paying so dearly for with the MJL4281/4302. The MJL3281/1302 has roughly the same SOA as the MJ15003/4 up to its 15 amp limit and costs only $2.07 at Arrow. The 4281/4302 is a more rugged part if you plan to use it on stage, but in a home/studio application the 3281/1302 should be suitable.
I wish that I could help design the protection circuit. Thanks again for all the work you've shared with us Jens.
BTW, I have tried various resistors (not Holco) and have had no problems with the leads not fitting. A snug fit in spots, but a fit.
Hi Still4given,
You can assemble the rest of the PCB and all the mechanical & PSU components. It is only 12(times 2) resistors that need to be omitted from the current buy. Think about ordering up one of those resistor kits with 10 of each E24 value from 10R to 2M2 500mW 1% metal film. They are surprisingly cheap and will last you years. Although at your pace more like six weeks.
While you are waiting, download Leach's articles (version 1 thro' 4.5) on the low tim and read it several times until you understand what you are trying to achieve. These are an excellent instruction on designing an amp although you would benefit from a more basic grounding if your electronics is not quite up to speed.
Just slow down on the building and stop making us so envious.
You can assemble the rest of the PCB and all the mechanical & PSU components. It is only 12(times 2) resistors that need to be omitted from the current buy. Think about ordering up one of those resistor kits with 10 of each E24 value from 10R to 2M2 500mW 1% metal film. They are surprisingly cheap and will last you years. Although at your pace more like six weeks.
While you are waiting, download Leach's articles (version 1 thro' 4.5) on the low tim and read it several times until you understand what you are trying to achieve. These are an excellent instruction on designing an amp although you would benefit from a more basic grounding if your electronics is not quite up to speed.
Just slow down on the building and stop making us so envious.
- Status
- Not open for further replies.