Terry,
That's exactly what I'm saying...
I used trimpots with legs about 1/4" long. all one has to do is route the correct leg to the correct hole. Just like using the mpsa transistors instead of 2sc.
Now some people won't like the 1 turn trimmers I'm using, and so may be committed to using the trace cutting method, but them's the breaks...
The pots cannot sit flat on the board after this, so I have to use some care adjusting them, and will ultimately 'glue' them to secure the position, but for now, no problem.
Stuart
That's exactly what I'm saying...
I used trimpots with legs about 1/4" long. all one has to do is route the correct leg to the correct hole. Just like using the mpsa transistors instead of 2sc.
Now some people won't like the 1 turn trimmers I'm using, and so may be committed to using the trace cutting method, but them's the breaks...
The pots cannot sit flat on the board after this, so I have to use some care adjusting them, and will ultimately 'glue' them to secure the position, but for now, no problem.
Stuart
Mmmm, Terry, do you read the pdf's you post?
Bourns are nice, even shows a picture of legs 1,2 and 3 and a schematic drawing with leg numbers.
Sticking legs 1 and 3 in one hole may have a peculiar effect !
Funny thing about these trimpots is that they are all listed as good for 1000 hours and 200 turns.
I never realised that i have to exchange trimpots every year !
Bourns are nice, even shows a picture of legs 1,2 and 3 and a schematic drawing with leg numbers.
Sticking legs 1 and 3 in one hole may have a peculiar effect !
Funny thing about these trimpots is that they are all listed as good for 1000 hours and 200 turns.
I never realised that i have to exchange trimpots every year !
legs...
...are quite sort on those guys. They are very nice pots, but I'm not sure you'll have enough room to get everything where you need it..
I can show you quite easily where the 3 pins have to end up, I don't have any batteries for my camera right now or I'd test it's macro feature to show you the boards I've done...
Looking at the board from above, input towards you, drivers on the edge farthest from you, the wiper has to go in the leftmost hole, the 'middle' hole gets the 'anticlockwise' track end, and so on...
I just checked, the legs on the trimmers I have are nearly 1/2" long, they end up about 1/4" tall after installation...
I got mine from the local electronics store, it's an NTE part...
http://www.nteinc.com/pot_web/75h.html
HTH
Stuart
...are quite sort on those guys. They are very nice pots, but I'm not sure you'll have enough room to get everything where you need it..
I can show you quite easily where the 3 pins have to end up, I don't have any batteries for my camera right now or I'd test it's macro feature to show you the boards I've done...
Looking at the board from above, input towards you, drivers on the edge farthest from you, the wiper has to go in the leftmost hole, the 'middle' hole gets the 'anticlockwise' track end, and so on...
I just checked, the legs on the trimmers I have are nearly 1/2" long, they end up about 1/4" tall after installation...
I got mine from the local electronics store, it's an NTE part...
http://www.nteinc.com/pot_web/75h.html
HTH
Stuart
jacco vermeulen said:Mmmm, Terry, do you read the pdf's you post?
Bourns are nice, even shows a picture of legs 1,2 and 3 and a schematic drawing with leg numbers.
Sticking legs 1 and 3 in one hole may have a peculiar effect !
Funny thing about these trimpots is that they are all listed as good for 1000 hours and 200 turns.
I never realized that i have to exchange trimpots every year !
Hi jacco,
Yes I looked it over but I don't really understand how the pot works in the circuit so I didn't know which leg needed to go where. Sorry for all the dumb questions.
Hi Stuart,
Thanks for the explanation. You are probably right about the legs being too short. I should probably just cut and jumper the traces.
Thanks Terry
bias trimpot
I used a standard bourns and cut the traces; but I'm still not sure on bias adjustment.
How about this - the bias trimpot sits between Q108 and the bias pad. If the resistance between these two points is increased do you get more or less bias? This way I can simply measure it and set it to the lowest (bias) value.
Terry- sorry for making a joke at your expense; i'll behave from now on.
Jacco- that table is a DIY item; good luck with yours. Maybe I'll put a page up with the DIY woodworking stuff.
I used a standard bourns and cut the traces; but I'm still not sure on bias adjustment.
How about this - the bias trimpot sits between Q108 and the bias pad. If the resistance between these two points is increased do you get more or less bias? This way I can simply measure it and set it to the lowest (bias) value.
Terry- sorry for making a joke at your expense; i'll behave from now on.
Jacco- that table is a DIY item; good luck with yours. Maybe I'll put a page up with the DIY woodworking stuff.
I read in earlyer posts that a lower resistance gives more bias current. Higher resistance gives a lower bias current.
bias...
As I mentioned before, there are few different ways of determining the 'correct' way to turn the pot...
Higher resistance == lower bias, arrange connections and adjust pot accordingly...
or disconnect output stage, measure the voltage from the drivers that would be fed to the bases of the output transistors, lower voltage == lower bias, less than 1.2v from one base to the other == no bias, expect ~2.0v as the bias reaches the correct point...
There are probably other methods as well, but these are easy to understand and implement...
Stuart
As I mentioned before, there are few different ways of determining the 'correct' way to turn the pot...
Higher resistance == lower bias, arrange connections and adjust pot accordingly...
or disconnect output stage, measure the voltage from the drivers that would be fed to the bases of the output transistors, lower voltage == lower bias, less than 1.2v from one base to the other == no bias, expect ~2.0v as the bias reaches the correct point...
There are probably other methods as well, but these are easy to understand and implement...
Stuart
Hi Stuart,
KSA50 is my starter being a good quality pcb with all of the topology that I really need.
Some or none of the extra bells and whistles will be assessed on sound quality and/or reliability.
Jacco,
Separate supply to input & VAS is a given, the real decision is how many volts above output PSU dc and whether to regulate.
I will NOT regulate the output PSU.
KSA50 is my starter being a good quality pcb with all of the topology that I really need.
Some or none of the extra bells and whistles will be assessed on sound quality and/or reliability.
Jacco,
Separate supply to input & VAS is a given, the real decision is how many volts above output PSU dc and whether to regulate.
I will NOT regulate the output PSU.
Normal and Low Bias
I am thinking about a switch to select "normal" (high) and "low" bias using the pads on the board.
Looks like the pads give you the option of switching in R144 in parallel to the bias resistor R126. This is only able to decrease overall resistance-- increasing the bias.
So for "normal" (high bias) mode, you would use both resistors in parallel, and for "low bias" mode you would break the connection between R144 and R126, increasing resistance and lowering bias.
And better yet, R126 is always in the circuit, so it should not matter whether you use a make-before-break switch to switch this in either direction.
Have I got this right?
I am thinking about a switch to select "normal" (high) and "low" bias using the pads on the board.
Looks like the pads give you the option of switching in R144 in parallel to the bias resistor R126. This is only able to decrease overall resistance-- increasing the bias.
So for "normal" (high bias) mode, you would use both resistors in parallel, and for "low bias" mode you would break the connection between R144 and R126, increasing resistance and lowering bias.
And better yet, R126 is always in the circuit, so it should not matter whether you use a make-before-break switch to switch this in either direction.
Have I got this right?
Re: bias...
If you are not sure about the bias and need to begin at the low bias point and are not sure which way to turn the trimpot to reduce bias, here is a foolproof trick that has helped me over the years.
Disconnect the driver's emitter to the output's base (saving the outputs from being driven). Then measure the voltage across the B and E of the driver, less than 0.6vdc and you are ok to go ahead with initial biasing, (This is class AB territory) above 0.6vdc and you are in high bias/ Class-A territory and you need to lower the bias to 0.5 before you connect the driver to the output stage. This method is a quick test of verification and not extremey accurate but very practical if you are not sure which way the trimpot should be.
Class-AB amps need to have about 0.50 to 0.55 vdc across the BE of the output devices as a quick test.
Another quick test I do it use a 100-200 watts lamp in series with the amplifier, then turn the amp on and set the trimpots such that the lamp goes down to a dull amber... this is low bias setting. Then remove the lamp and set the bias correctly using the voltage readings off the OP device's emmitter resistors as others have explained.
K-
Stuart Easson said:
If you are not sure about the bias and need to begin at the low bias point and are not sure which way to turn the trimpot to reduce bias, here is a foolproof trick that has helped me over the years.
Disconnect the driver's emitter to the output's base (saving the outputs from being driven). Then measure the voltage across the B and E of the driver, less than 0.6vdc and you are ok to go ahead with initial biasing, (This is class AB territory) above 0.6vdc and you are in high bias/ Class-A territory and you need to lower the bias to 0.5 before you connect the driver to the output stage. This method is a quick test of verification and not extremey accurate but very practical if you are not sure which way the trimpot should be.
Class-AB amps need to have about 0.50 to 0.55 vdc across the BE of the output devices as a quick test.
Another quick test I do it use a 100-200 watts lamp in series with the amplifier, then turn the amp on and set the trimpots such that the lamp goes down to a dull amber... this is low bias setting. Then remove the lamp and set the bias correctly using the voltage readings off the OP device's emmitter resistors as others have explained.
K-
Hi Guys,
I've only recently joined the forum and have been following the discussions with interest. I'm looking at options on components and this has thrown up a small problem. Lloyd Macleans excellent information and schematics for the KSA50 show MJ15004 transistors in the positive half of the output stage and these are drawn as NPN units. However, all the sources I can get these from state that they are PNP items. Am I being stupid here and missing something or is it safe to assume there is a typo that needs correcting.
Cheers,
valvemaniac.
I've only recently joined the forum and have been following the discussions with interest. I'm looking at options on components and this has thrown up a small problem. Lloyd Macleans excellent information and schematics for the KSA50 show MJ15004 transistors in the positive half of the output stage and these are drawn as NPN units. However, all the sources I can get these from state that they are PNP items. Am I being stupid here and missing something or is it safe to assume there is a typo that needs correcting.
Cheers,
valvemaniac.
xistors
Well since I usally ask the questions around here lets see if I can answer yours for a change.
I've got a whole tray of MJ15004 sitting in front of me. They are PNP. MJ15003 are NPN. OK, that was kind of stupid you can't really tell from looking at them.
Datasheet here: http://www.onsemi.com/pub/Collateral/MJ15003-D.PDF
Well since I usally ask the questions around here lets see if I can answer yours for a change.
I've got a whole tray of MJ15004 sitting in front of me. They are PNP. MJ15003 are NPN. OK, that was kind of stupid you can't really tell from looking at them.
Datasheet here: http://www.onsemi.com/pub/Collateral/MJ15003-D.PDF
Yes Igreen, you are right about your next statement.
"And better yet, R126 is always in the circuit, so it should not matter whether you use a make-before-break switch to switch this in either direction".
See me earlyer post about a switch for high and low bias, making paralell of this manual switch one temperatureswitch for each channel, so by high temp, a relais switches from high bias to low bias. Automaticly. The status of the bias is indicated with 2 leds on the front of the amp. See schematic.
"And better yet, R126 is always in the circuit, so it should not matter whether you use a make-before-break switch to switch this in either direction".
See me earlyer post about a switch for high and low bias, making paralell of this manual switch one temperatureswitch for each channel, so by high temp, a relais switches from high bias to low bias. Automaticly. The status of the bias is indicated with 2 leds on the front of the amp. See schematic.
valvemaniac said:
You may also take a look at the single stage VAS of the KSA50 MK1 on Mr. MacLeans 's homepage.
The Mark1 had MDS21/MDS60, replacements would be MJE340/350, or other obsolete devices like the MJE3440.
Frequency compensation in the MK1 is 39 pF, not 390 pF.
Considering the giant step forward from the MK1 to the MK2, i'd question the 390pF value.
(you fellows be glad Pavel didn't hop the boat to the US.)
1 out of 2 aint bad.
Well, I got the proto boards from Mark yesterday, and I am in the process of building them up, I just need to get some more 47K resistors.
They are very nice, and apart from the holes issue that we know about, they go together really well. However,there are several minor things I would change, but nothing that will affect functionality, just little things that I noticed that no-one else will even care about, even if they see them.
Ah, the pains of being an artist... 🙂
They are very nice, and apart from the holes issue that we know about, they go together really well. However,there are several minor things I would change, but nothing that will affect functionality, just little things that I noticed that no-one else will even care about, even if they see them.
Ah, the pains of being an artist... 🙂
jacco vermeulen said:
What is it that you don´t agree with jacco?
I think my Zapsolutes sounds a little more relaxed/resolved and refined (the 3 R 🙂 ) running only one pair compared to four pairs.
One very good Swedish audio designer tested a single pair Zapsolute side by side with a KSA80 and the KSA80 was a clear looser in everything except the bass (where we now can use class D). The speaker during thes tests used a Raven ribbon, C79 mids and some TC UH woofers.
I would buy a Zapsolute instead of fippling around with a Krell clone any day. GamuT audio also agrees with LC audio regarding paralelled outputs or not.
Single outputs, class A no NFB for mids/highs and parallelled outputs with feedback for lows maybe.. or maybe class D for lows.
/Peter
Attention to detail is what separates the proffesionals from the amateurs Al. Care to tell us what those imperfections are you want to change?