Krell transformer information:
Hi all Krell builders, especially Gabor:
Here are some data for the original Krell transformers
Secondary windings:
KMA 80, KMA 160:
2 x 0-32-45V (800VA?)
1 x 11-0-11V
1 x 70-0-70V
+ screen
-----------------------------
KSA 200, KSA 400
2 x 0-55-70V (1500VA?)
1 x 11-0-11V
1 x 70-0-70V
+ screen
-----------------------------
KSA 100 (mk-I + mk-II)
2 x 0-41V (900VA)
1 x 0-12V
1 x 62-0-62V
+ screen
-----------------------------
All primaries was 2 x 0-110-120V
Regarding VA ratings: the only model I can confirm is the mk-I, + mk-II models, the other models is based on a qualified only guessing.
If using only one xformer for a stereo amp you can use a unit with two or three volts lower rail voltage windings than the original specs (for the VA-rated windings) due to better regulation for a bigger transformer.
Thats why Mark suggested 38-0-38V for the mk-II stereo amp (if someone was wandering)
Why mk-I and mk-II had that 62-0-62V winding I cant figure out. Anybody here who knows (it was not for the LTP-predriver stage)?
Regards
Hi all Krell builders, especially Gabor:
Here are some data for the original Krell transformers
Secondary windings:
KMA 80, KMA 160:
2 x 0-32-45V (800VA?)
1 x 11-0-11V
1 x 70-0-70V
+ screen
-----------------------------
KSA 200, KSA 400
2 x 0-55-70V (1500VA?)
1 x 11-0-11V
1 x 70-0-70V
+ screen
-----------------------------
KSA 100 (mk-I + mk-II)
2 x 0-41V (900VA)
1 x 0-12V
1 x 62-0-62V
+ screen
-----------------------------
All primaries was 2 x 0-110-120V
Regarding VA ratings: the only model I can confirm is the mk-I, + mk-II models, the other models is based on a qualified only guessing.
If using only one xformer for a stereo amp you can use a unit with two or three volts lower rail voltage windings than the original specs (for the VA-rated windings) due to better regulation for a bigger transformer.
Thats why Mark suggested 38-0-38V for the mk-II stereo amp (if someone was wandering)
Why mk-I and mk-II had that 62-0-62V winding I cant figure out. Anybody here who knows (it was not for the LTP-predriver stage)?
Regards
Hi,
The ksa100 41Vac seems high. This would give Vdc after taking out the Iq=2.6A of about +-58Vdc. I thought the rails were at +-52Vdc.
Now I don't know anything!
maybe they bought in the transformers with the intention of a regulated front end. Then changed their mind to save money (who Krell??).
The ksa100 41Vac seems high. This would give Vdc after taking out the Iq=2.6A of about +-58Vdc. I thought the rails were at +-52Vdc.
I was planning on using 37-0-37Vac, based on +-52Vdc.
Now I don't know anything!
Yes you are right jacco
I was to late for to edit my post. Ofcourse it should be 2 x 0-100-120 on all primaries.
sorry for that
Andrew, your question is relevant. If for example the ACV was measured under heavy load then this is the case: 41 -1.5 or 2V in rectifier, then 39V x SQRT2 = 55V DC rails. This is 3VDC more than what is shown in the schematics. Could there have been a larger forward drop in the rectifier at that time (in the -80´s) than what we calculate for today (more efficient diodes for today)?
Maybe Krell have specified a minimum DC rail Voltage and a maximum ACV for the KSA 100 for to keep the specs and power diss. at right levels ?
Hummmmm?????
Regards
I was to late for to edit my post. Ofcourse it should be 2 x 0-100-120 on all primaries.
sorry for that
Andrew, your question is relevant. If for example the ACV was measured under heavy load then this is the case: 41 -1.5 or 2V in rectifier, then 39V x SQRT2 = 55V DC rails. This is 3VDC more than what is shown in the schematics. Could there have been a larger forward drop in the rectifier at that time (in the -80´s) than what we calculate for today (more efficient diodes for today)?
Maybe Krell have specified a minimum DC rail Voltage and a maximum ACV for the KSA 100 for to keep the specs and power diss. at right levels ?
Hummmmm?????
Regards
Flodstroem said:
Exactly, minimum is the key word.
The secondary output voltage varies, depending on the mains voltage level.
Also a reason why tested power ratings of US amplifiers of that era were often much higher than specified.
The toroid has 100Vac and 120Vac primaries, the entrance voltage is 120Vac in the schematic.
But US main voltage ranged from 110Vac, 115Vac, to 120Vac, depending on the location.
At 120Vac mains voltage on the 120Vac primaries you'll have 41Vac on the secondaries.
At 110Vac, secondaries voltage is 37.6 Vac. Rectified that makes around 52Vdc.
Also a reason why the big cans in the Krell were 75V types.
63V electrolytics would suffice at 110Vac, but 120Vac requires 75V electrolytics.
This is old familiar stuff, for anyone who fiddled power amplifiers in the previous millenium.
And the schematic shows your 62Vac winding answer.
I guess most of you have never seen the huge industrial relays that were used by Krell back then.
And that leads to questioning why anyone would want to use highly expensive custom-order toroids.
Only half of the auxilliary secondary windings were used on the Krell, and for a Krell replica you're likely not the use those either.
Leaves you with a toroid that has 55-57Vdc rails at 2K mains voltage levels.
Pierre, you have 220Vac mains where you reside ?
jacco:
Ah-ha yes I was thinking of such a thing, and, yes I have been working with this type of relays in the 70´s and 80´s
.
jacco,
One idea could be to re-customize the Krell original transformer to better suit our needs for today, here is a suggestion:
Secondaries
1) 0-38V @ 450VA
2) 0-38V @ 450VA
3) 0-20V ca 5-10VA (for LTP and pre-driver)
4) 0-20V ca 5-10VA (for LTP and pre-driver)
5) 0-14V 10VA optional, for logic
+ shield
Primaries
1) 0-110-115-120
2) 0-110-115-120
Will cover 110, 115, 120, US mains voltage and 220, 230 and 240 rest of the worlds mains voltage (most of it)
The xformer total VA will be 900+ (for a single channel/ or mono-block).
DCV should then be ±52VDC minimum rails + ±80VDC for a regulated ±39V for LTP plus ± 60-70V for pre-driver/voltage amp (the KMA-style
).
the optional 14VAC gives 18-19VDC for the 12V reg IC could drive fans, protection circuit, power on delay etc... if not using a separated small xformer.
If your plans are to use only one xformer in a stereo amp just double the VA ratings.
Regards
Ah-ha yes I was thinking of such a thing, and, yes I have been working with this type of relays in the 70´s and 80´s
.jacco,
One idea could be to re-customize the Krell original transformer to better suit our needs for today, here is a suggestion:
Secondaries
1) 0-38V @ 450VA
2) 0-38V @ 450VA
3) 0-20V ca 5-10VA (for LTP and pre-driver)
4) 0-20V ca 5-10VA (for LTP and pre-driver)
5) 0-14V 10VA optional, for logic
+ shield
Primaries
1) 0-110-115-120
2) 0-110-115-120
Will cover 110, 115, 120, US mains voltage and 220, 230 and 240 rest of the worlds mains voltage (most of it)
The xformer total VA will be 900+ (for a single channel/ or mono-block).
DCV should then be ±52VDC minimum rails + ±80VDC for a regulated ±39V for LTP plus ± 60-70V for pre-driver/voltage amp (the KMA-style
).the optional 14VAC gives 18-19VDC for the 12V reg IC could drive fans, protection circuit, power on delay etc... if not using a separated small xformer.
If your plans are to use only one xformer in a stereo amp just double the VA ratings.
Regards
I have contacted the guy I use for custom transformers, he has come back with the following quotation.
1200VA core,
mechanically stabilised, primary windings overwound to prevent movement.
Primary either 110 or 230V
Secondary 1: 45-0-45 @ 1KVA
Secondary 2: 0-15 @ 250 mA
Secondary 3: 0-15 @ 250 mA
Slight adjustment of voltage values (in case I got them wrong) will not effect cost.
Cost £70.00
Carriage £12.50 for up to 4 transformers within UK.
1200VA core,
mechanically stabilised, primary windings overwound to prevent movement.
Primary either 110 or 230V
Secondary 1: 45-0-45 @ 1KVA
Secondary 2: 0-15 @ 250 mA
Secondary 3: 0-15 @ 250 mA
Slight adjustment of voltage values (in case I got them wrong) will not effect cost.
Cost £70.00
Carriage £12.50 for up to 4 transformers within UK.
KSA 100 M2 upgrade feedback
Hi
Some time back the amplifier driver stage, output devices, the internal wires between the pcb and output devices were changed to Van den Hul CS16, the Roederstein pcb caps replaced with one Black Gate cap per board and the mains input changed from 115 volt to 240 volt. My expectation was that all these upgrades would contribute to an overall improvement in sound quality.
Listening to the amp after the changes my initial impressions was that it had lost the Krell characteristics. The soundstage was skimpy, I was hearing two speakers, vocal pieces sounded shrill and lifeless and the extreme lower bass sounded loose. Apart from a volume capacity there was nothing to suggest that this amp had the finesse it is known for.
My initial thoughts re the causes were that it could be:
a) The BG caps that still needed some more time to burn-in?
b) More accurate output transistor matching was required?
c) Or to revert back to the slower original driver transistors on the pcb?
d) Drop in voltage when using 240 transformer setting?
Yesterday after my Sunday afternoon nap I woke up with the idea that I need to remove the tight cable ties that held the yellow Van Den Hull cable in place. Interestingly once the cable ties were removed there was a significant improvement in the soundstage solidity and inner detail.
The lesson here seems to be that there is obviously more subtle issues at stake here. The Van Den Hul wire obviously did not like to be “crimped” or to be so close to the power wiring of the output transistors. This is possibly a hint to everyone that care needs to be taken in the wire layout if optimum results is a criteria- especially if you have no idea of how a Krell should sound.
Jozua
Hi
Some time back the amplifier driver stage, output devices, the internal wires between the pcb and output devices were changed to Van den Hul CS16, the Roederstein pcb caps replaced with one Black Gate cap per board and the mains input changed from 115 volt to 240 volt. My expectation was that all these upgrades would contribute to an overall improvement in sound quality.
Listening to the amp after the changes my initial impressions was that it had lost the Krell characteristics. The soundstage was skimpy, I was hearing two speakers, vocal pieces sounded shrill and lifeless and the extreme lower bass sounded loose. Apart from a volume capacity there was nothing to suggest that this amp had the finesse it is known for.
My initial thoughts re the causes were that it could be:
a) The BG caps that still needed some more time to burn-in?
b) More accurate output transistor matching was required?
c) Or to revert back to the slower original driver transistors on the pcb?
d) Drop in voltage when using 240 transformer setting?
Yesterday after my Sunday afternoon nap I woke up with the idea that I need to remove the tight cable ties that held the yellow Van Den Hull cable in place. Interestingly once the cable ties were removed there was a significant improvement in the soundstage solidity and inner detail.
The lesson here seems to be that there is obviously more subtle issues at stake here. The Van Den Hul wire obviously did not like to be “crimped” or to be so close to the power wiring of the output transistors. This is possibly a hint to everyone that care needs to be taken in the wire layout if optimum results is a criteria- especially if you have no idea of how a Krell should sound.
Jozua
Attachments
The Omron LY4 is able to switch currents of up to 40 amps with 4 contacts in parallel and has a life expectancy that can be as high as 1 Milllion switching cycles.
But with delay/switching times of over 25mS it is/was not a sota relay.
The 62Vac winding may have enabled an easy timing delay for a 24Vdc relay, but what a spenderish way.
I suppose it's like with the KMA servo, the d'A brothers must have had a great disliking of integrated critters in their amp products.
Here those relays used to be a source of trouble for KSA models, maybe because of the climate. As active protection they were of little to no use imo, way to slow. Some KSA owners even exchanged the relays and protection circuitry.
Some time ago i bought a box of relays for 1 Euro the piece from someone offering them on a diyA thread, enough to keep switching me on/off for decades after i've gone 6ft under.
Their current wholesale listing price is 10-15 times as much, in the late 80s i paid in excess of $10/pc for them.
Those have data that are 5-10 times better than the LY4s, with 2 in parallel they manage 40 amps continuous peak levels as well and in contrast to the Omrons they are fully sealed and innert gas filled.
I may not be true audiofilo grade by favoring some sort of DC and peak current output protection.
Anyone who does not object to using high quality relays is better off with a small additional transformer.
That would also be an easy way to go regulated for the front end and driver stages.
The idea of the KSA-100 was to double the output power into half the load impedance for a minimum of 400 watts continuous power in 2 Ohms, and with a minimum power of 100 watts class A in 8 Ohms.
If someone desires more than the 400/2, then constructing a KMA is more logical.
Which cables were tied together ?
But with delay/switching times of over 25mS it is/was not a sota relay.
The 62Vac winding may have enabled an easy timing delay for a 24Vdc relay, but what a spenderish way.
I suppose it's like with the KMA servo, the d'A brothers must have had a great disliking of integrated critters in their amp products.
Here those relays used to be a source of trouble for KSA models, maybe because of the climate. As active protection they were of little to no use imo, way to slow. Some KSA owners even exchanged the relays and protection circuitry.
Some time ago i bought a box of relays for 1 Euro the piece from someone offering them on a diyA thread, enough to keep switching me on/off for decades after i've gone 6ft under.
Their current wholesale listing price is 10-15 times as much, in the late 80s i paid in excess of $10/pc for them.
Those have data that are 5-10 times better than the LY4s, with 2 in parallel they manage 40 amps continuous peak levels as well and in contrast to the Omrons they are fully sealed and innert gas filled.
I may not be true audiofilo grade by favoring some sort of DC and peak current output protection.
Anyone who does not object to using high quality relays is better off with a small additional transformer.
That would also be an easy way to go regulated for the front end and driver stages.
The idea of the KSA-100 was to double the output power into half the load impedance for a minimum of 400 watts continuous power in 2 Ohms, and with a minimum power of 100 watts class A in 8 Ohms.
If someone desires more than the 400/2, then constructing a KMA is more logical.
Which cables were tied together ?
The cable ties were used (just like the original) to tie the interconnect to the DC power fed to the rear transistors of the blocks.
I've thought for a brief while about possible issues but the original wire were super-thin and thus in much closer proximity than the very thick vdH, plus the vdH coax is balanced with the shield tied to ground at one end and the signal and return connected to the two internal conductors.
I've thought for a brief while about possible issues but the original wire were super-thin and thus in much closer proximity than the very thick vdH, plus the vdH coax is balanced with the shield tied to ground at one end and the signal and return connected to the two internal conductors.
Hi cantskienuf,
The new Wiki is located here:
http://www.diyaudio.com/wiki/index.php?page=Add+your+names+join+the+group+buy
Regards
The new Wiki is located here:
http://www.diyaudio.com/wiki/index.php?page=Add+your+names+join+the+group+buy
Regards