News: Trouble logging in? Log in from the login page.
September 22, 2019, 08:17:20 AM

Author Topic: Output impedance switching  (Read 411 times)

0 Members and 1 Guest are viewing this topic.

Offline Raymond P.

  • Full Member
  • ***
  • Posts: 97
Output impedance switching
« on: May 21, 2019, 10:44:08 AM »
I have two pairs of speakers, one with an 8-ohm nominal impedance and another 4-ohms, that I'd like to alternately drive with my Kaiju. A recent topic here gave me the idea of switching the output transformer impedance between 4 and 8 ohms conveniently. Reviewing the schematics and instructions, it looks like I need a three pole, double throw switch for each OT to select between the two configurations. Has anyone done this? Anyone see any issues?


One aspect I'm not sure about is where to mount the two 3pdt switches. I do not want to drill into the top plate or mount on the wood base. I could remove the trim pots (I don't use them) and use those holes, but I can't find any switches that'll fit. It seems those holes are 9-10mm in diameter, and these switches require either a 6mm or 12mm opening. Any suggestions?


Thanks.
Raymond P.

Offline Deke609

  • Sr. Member
  • ****
  • Posts: 582
Re: Output impedance switching
« Reply #1 on: May 25, 2019, 05:47:48 PM »
I don't have a Kaiju, but I added impedance switch boards to my Stereomour II using PJ's schematic: http://bottlehead.com/smf/index.php?topic=11130.0

If the Kaiju OTs have the same winding configuration as the SII OTs, you should only need 2 DPDT switches, assuming you do not intend to add a balanced/unbalanced switch.

I built the impedance switches on perforated proto-board using slide switches and mounted the boards on long standoffs attached to the OT mounting screws. They sit on top of the OTs. I recommend getting low profile switches and measuring the height of the OTs to ensure that there is clearance between the switches and the surface upon which the amp sits.

cheers,

Derek
Derek
______

Roon Intel NUC ->  Yggdrasil DAC -> BeePre (w/ BeeQuiet and EML 300B)  -> Stereomour IIs (2A3 [EML 2A3 Mesh] and 45 Conversion [EML 45B] - both fully upgraded)  -> Audeze LCD4

Online Paul Birkeland

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 11317
Re: Output impedance switching
« Reply #2 on: May 26, 2019, 05:13:59 AM »
I believe that this will work for the OT-5, but I would wait until PJ gets home and confirms 100% for sure before implementing it.

Edit: looking at the Kaiju manual, the output transformer terminal number does not appear to be consistent between the Paramount and Kaiju, though this drawing could probably be renumbered to remedy that.

-PB
« Last Edit: May 26, 2019, 07:37:37 AM by Paul Birkeland »
Paul "PB" Birkeland

Bottlehead Grunt & The Repro Man

Offline Raymond P.

  • Full Member
  • ***
  • Posts: 97
Re: Output impedance switching
« Reply #3 on: May 26, 2019, 07:04:03 AM »
Thanks, Derek for the reference to the Stereomour II topic and the switch mounting suggestions.

I realized I make an embarrassing mistake when I said that I needed a 3pdt switch per OT. I was actually looking at the Stereomour II manual (I have the kit on order). Looking at the Kaiju manual, I really need one 4pdt switch per OT to make identical sets of connections as indicated in the wiring instructions for 4 and 8 ohms. From what I can gather, the Stereomour II and Kaiju OTs have the same secondary winding scheme: two sections with a center tap (as shown in PB's diagram above.) The kits differ in how the OT's sections are connected in series/parallel to get the different output impedances, hence 3pdt vs 4pdt. I'd love to know the reason why that's so. But going by schematics along, one kind of switch, any kind, should work for 4 and 8 ohm switching for both the SII and Kaiju, if making identical connections per the instructions isn't necessary.

I only need to switch between 4 and 8 ohms, so in any case I should only need one switch per OT, which is good because I there are two holes in the Kaiju top plate I can repurpose. (I've decided I'd prefer to have the switches easily visible and accessible.)
« Last Edit: May 26, 2019, 07:07:25 AM by Raymond P. »
Raymond P.

Offline Raymond P.

  • Full Member
  • ***
  • Posts: 97
Re: Output impedance switching
« Reply #4 on: May 26, 2019, 09:05:41 AM »
I believe that this will work for the OT-5, but I would wait until PJ gets home and confirms 100% for sure before implementing it.

Edit: looking at the Kaiju manual, the output transformer terminal number does not appear to be consistent between the Paramount and Kaiju, though this drawing could probably be renumbered to remedy that.

-PB


This circuit is great because that means I would only need one 2pdt switch per OT since I only want the 4 and 8 ohm configurations. In trying to understand this, I traced the 8 ohm configuration and got this simplified circuit. Is this correct, because that would give a 10 ohm impedance?


« Last Edit: May 26, 2019, 09:10:16 AM by Raymond P. »
Raymond P.

Online Paul Birkeland

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 11317
Re: Output impedance switching
« Reply #5 on: May 26, 2019, 09:50:30 AM »
There isn't a separated 2 ohm winding on the OT5, it's a tap on one of the 4 ohm windings.  Two 4 ohm windings in series gives 16 ohms.  If you take a 4 ohm winding and put it in series with a portion of the other 4 ohm winding, you get to 8 ohms. 

I started drawing out impedance switching for all four taps, but ran out of brain power and decided to eat lunch instead.
Paul "PB" Birkeland

Bottlehead Grunt & The Repro Man

Offline Raymond P.

  • Full Member
  • ***
  • Posts: 97
Re: Output impedance switching
« Reply #6 on: May 26, 2019, 10:00:53 AM »
There isn't a separated 2 ohm winding on the OT5, it's a tap on one of the 4 ohm windings.  Two 4 ohm windings in series gives 16 ohms.  If you take a 4 ohm winding and put it in series with a portion of the other 4 ohm winding, you get to 8 ohms. 

I started drawing out impedance switching for all four taps, but ran out of brain power and decided to eat lunch instead.
Hi, Paul.
I got the 2 ohm winding in my simplified diagram from two parallel windings: the 1-6 in parallel with the 3-8 (when the switches are in the 8 ohm position), unless I'm just not reading the diagram incorrectly.
« Last Edit: May 26, 2019, 10:02:41 AM by Raymond P. »
Raymond P.

Online Paul Birkeland

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 11317
Re: Output impedance switching
« Reply #7 on: May 26, 2019, 10:08:47 AM »
Once you've put the two 4 ohm windings in series, there is no more copper to put in series with those.
Paul "PB" Birkeland

Bottlehead Grunt & The Repro Man

Offline Raymond P.

  • Full Member
  • ***
  • Posts: 97
Re: Output impedance switching
« Reply #8 on: May 26, 2019, 10:16:18 AM »
Hi Paul.
I'd like to confirm an assumption I'm making. Are the 1-6, 6-2, 7-3, and 3-8 windings in your diagram each 4 ohms? Thx.
Raymond P.

Online Paul Birkeland

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 11317
Re: Output impedance switching
« Reply #9 on: May 26, 2019, 10:20:35 AM »
Hi Paul.
I'd like to confirm an assumption I'm making. Are the 1-6, 6-2, 7-3, and 3-8 windings in your diagram each 4 ohms?
No. 
Paul "PB" Birkeland

Bottlehead Grunt & The Repro Man

Offline Paul Joppa

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 4879
Re: Output impedance switching
« Reply #10 on: May 27, 2019, 12:56:31 PM »
The two transformers (OT-2 Stereomour/SEX) and OT-5 (Kaiju) have different secondaries:

1) OT-2 has one secondary wound in the reverse direction. This allowed a sneaky trick of wiring to minimize leakage inductance. In practice, the "sneaky trick" turned out to make it very confusing to figure out the wiring ...  :^)

2) For OT-5, I used a different approach to the same issue, which as a side advantage gave more exact turns ratio and equal losses for the different impedances. It does NOT use any reverse windings.

By the time Kaiju came out, the OT-2 switch board had been discontinued, so I never drew up a switch circuit for OT-5.

I'll just note that rotary switches are readily available in 4-pole, three-position form - it's what we use for source selector switching, for example. You can probably set one up for any three of the four possible impedances. And they should fit the gain trim potentiometer holes as proposed.
Paul Joppa

Offline Raymond P.

  • Full Member
  • ***
  • Posts: 97
Re: Output impedance switching
« Reply #11 on: May 27, 2019, 07:02:07 PM »
The two transformers (OT-2 Stereomour/SEX) and OT-5 (Kaiju) have different secondaries:

1) OT-2 has one secondary wound in the reverse direction. This allowed a sneaky trick of wiring to minimize leakage inductance. In practice, the "sneaky trick" turned out to make it very confusing to figure out the wiring ...  :^)

2) For OT-5, I used a different approach to the same issue, which as a side advantage gave more exact turns ratio and equal losses for the different impedances. It does NOT use any reverse windings.

Thanks. That explains the connection differences between the two that's not apparent from just schematics.
Quote
I'll just note that rotary switches are readily available in 4-pole, three-position form - it's what we use for source selector switching, for example. You can probably set one up for any three of the four possible impedances. And they should fit the gain trim potentiometer holes as proposed.

Cool!
Raymond P.

Offline Raymond P.

  • Full Member
  • ***
  • Posts: 97
Re: Output impedance switching
« Reply #12 on: May 28, 2019, 10:45:01 AM »
I was originally planning on switching between 4 and 8 ohm outputs using a toggle switch per OT, but the recommendation of using rotary switches makes it feasible to have more than 2 throws, so I've decided to have the ability to switch between all 4 output impedances. I came up with a circuit and attached the schematic for it in case anyone else is interested. I have not implemented it yet, and it may contain errors, so if anyone sees any issues, please let me know.

The switches in green are contained within one 4p4t rotary switch. The schematic shows one way to implement the switching. I believe there are other ways. I'd love to hear of better ways that would reduce the number of poles or pieces of wires to solder. TIA!


Edit: Here's a 4p4t rotary switch - https://smile.amazon.com/dp/B07JLQ38GP/ref=cm_sw_em_r_mt_dp_U_-aA7Cb9QR7NNS
« Last Edit: May 28, 2019, 10:59:43 AM by Raymond P. »
Raymond P.