Hoffman Amplifiers Tube Amplifier Forum
Amp Stuff => Tube Amp Building - Tweaks - Repairs => Topic started by: gfarina55 on November 30, 2021, 08:16:36 pm
-
I'm writing a BOM for an AB763 BF Deluxe Reverb build based on Robrob's layout. I'm not really interested in mods, just trying to build the best possible version of this classic amp. To those of you who have built this, anything you wish you'd known before you started? Willabe, you helped me make my 5F6A almost silent. Any benefit to moving the 2 x B+1 filter caps closer to the PT as you recommended for the 5F6A?
Also, I'm looking at Hammond transformers:
291BX Power
1760H Output
194A Choke
...but as usual I'm wondering about the higher wall voltage. Will the Hammonds give me an acceptable B+, or should I seek out something else?
Thanks!!
-
I've built several of these; they all sound glorious.
Doug Hoffman's turret board and build notes are absolutely first rate.
I don't think there is a benefit to keeping the filter caps close to the PT. In fact, Doug likes to tuck his in next to the turret board -- worth checking out, though I prefer the doghouse.
Hammond iron works great. Hawkusa has great prices and very fast shipping, and no I don't work for them or get a cut . . . No problem with voltage, though of course it will be higher than on the Fender schematic. I've used the iron you spec'd with 6V6 output tubes.
Take a look at Hoffman's build notes. Way more helpful than RR's layout (which of course is just the standard Fender layout).
-
Thx acheld, great info, I'll probably go w/the Hammonds. Regarding the layout, I'm a vintage freak and want to use the trad layout, or Robrob's version of that. I know Hoffman's solves some problems tho, so thx for mentioning.
-
Also, I'm looking at Hammond transformers:
291BX Power
1760H Output
194A Choke
...but as usual I'm wondering about the higher wall voltage. Will the Hammonds give me an acceptable B+, or should I seek out something else?
Thanks!!
You could use the 290CX (reissues) for lower voltage but i've had no problems with either, just set the bias properly.
-
I don't think there is a benefit to keeping the filter caps close to the PT. In fact, Doug likes to tuck his in next to the turret board -- worth checking out, though I prefer the doghouse.
His question about moving the B+ A node cap(s) is not in context.
It wasn't about keeping the 1st B+ A node filter cap(s) close to the PT. It was to be able to wire the PT's CT directly to the B+ A node cap(s) ground lead. Then run a wire from there over to the power amp chassis ground connection. He could have done this with all the filter caps in the dog house, but, this way he had less wires running back and forth from the dog house. It was Merlin's grounding.
Here's the thread with the layout drawings and details;
https://el34world.com/Forum/index.php?topic=27769.msg306841#msg306841 (https://el34world.com/Forum/index.php?topic=27769.msg306841#msg306841)
Any benefit to moving the 2 x B+1 filter caps closer to the PT as you recommended for the 5F6A?
See above.
You should be able to figure it out for yourself now. Go back over the thread. BF Deluxe Reverb amps are pretty quite as they are.
It's also going to depend on what chassis your going to use. A pre-punched clone or making it yourself?
Unless you have heard the 'roach' tremolo in that amp and love it, I would not use it. That is the least liked of the Fender trems, it's choppy, not very smooth.
I would either go with bias vary trem or Sluckey has a very smooth sounding trem he came up with. The problem with bias very trem can be trying to balance the strength of the trem with how hot the tubes are biased. To get strong(er) trem sometimes you have to bias the power tubes cooler. Not as bad with 6V6 tubes compared to more powerful tubes, like 6L6GB/GC tubes. Sluckey's trem doesn't have this problem.
And I would put in a mid range control in the second speaker jack hole, if using a pre-punched clone chassis.
I'd also make the 1st channel a lead channel and I'd put the verb on both channels. I think it's Robrob that has these mods?
-
Unless you have heard the 'roach' tremolo in that amp and love it, I would not use. That is the least liked of the Fender trems, it's choppy, not very smooth.
I would either go with bias vary trem or Sluckey has a very smooth sounding trem he came up with. The problem with bias very trem can be trying to balance the strength of the trem with how hot the tubes are biased. To get strong(er) trem sometimes you have to bias the power tubes cooler. Not as bad with 6V6 tubes compared to more powerful tubes, like 6L6GB/GC tubes. Sluckey's trem doesn't have this problem.
The Trem-O-Nator is very easy to incorporate into the original Fender AB763 layout. I helped Platefire do this to his Pro Reverb. You can find his thread in this forum if interested.
-
BF Deluxe Reverb amps are pretty quite as they are.
I meant to say, that is, if you split the power amp and preamp grounds. Pretty basic, easy to do.
-
Fantastic, thanks Slucky and Willabe. I will be using a pre-punched chassis, and I'll look into the Trem-O-Nater mod. Willabe, just to make sure I understand, you are saying:
"BF amps are pretty quiet.... if you split the power amp and preamp grounds." It appears to me they are already split in Robrob's layout, correct? Or are you suggesting moving them physically from the doghouse? (I'm trying to keep the layout as "vintage correct" as possible, so I'd like to keep the doghouse.)
-
"BF amps are pretty quiet.... if you split the power amp and preamp grounds." It appears to me they are already split in Robrob's layout, correct?
Yes, they are.
But the 1st B+ filter cap(s) ground is not wired directly to the PT's B+ CT.
Or are you suggesting moving them physically from the doghouse? (I'm trying to keep the layout as "vintage correct" as possible, so I'd like to keep the doghouse.)
We went all through this with your other build.
I'm not suggesting anything. I don't like to use the dog house. All my reasoning is in the other thread. I like Merlin and Kevin O'Connor's grounding.
-
The doghouse is fine. It has proven itself for over 60 years. But notice there are multiple ground wires leaving the doghouse and they need to connect to different points on the chassis. Rob's layout shows you how to correctly do this.
Of course, the Merlin disciples will argue that Merlin's improved grounding scheme is the only way to do it. But if you want to use a doghouse, then do it. I got your back... :icon_biggrin:
-
Ooops, I posted to the wrong thread.
-
OK, got it. Thx Willabe.
-
Quick question about grounding shielded wire: I'm running it from the grid stoppers (mounted on the input jacks) to the V1 and V2 grids as Fender did. If I'm using a Robinette-style pre-amp ground bus and my input jacks are not isolated, I should connect one end of the shield to the preamp ground bus (or to the input jack ground near the preamp ground bus termination point) and insulate the other end of the shielding from V1 and V2 grids, correct? Does anyone have a good photo of AB763 input jacks wired this way? Can't find anything helpful from google.
While searching for an answer to this question, I noticed the Hoffman AB763 layout (see below) grounds the preamp ground bus to the power amp side of the chassis instead of at the inputs. Just curious why? I thought the idea was to separate the pre- and power-amp grounds?
Thanks!!
-
Quick question regarding shielded wire on the inputs: If I'm using a Robinette-style pre-amp ground bus and my input jacks are not isolated, I should connect one end of the shield to the preamp ground bus (or to the input jack ground near the preamp ground bus termination point) and insulate the other end of the shielding from V1 and V2 grids, correct? Does anyone have a good photo of AB763 input jacks wired this way? Can't find anything helpful from google.
While searching for an answer to this question, I noticed the Hoffman AB763 layout (see below) grounds the preamp ground bus to the power amp side of the chassis instead of at the inputs. Just curious why? I thought the idea was to separate the pre- and power-amp grounds?
That is confusing. I wish Doug would move that ground symbol.
If you study Doug's layout you'll see exactly how he wired the shielded cables for the input jacks and the volume pots. Look closely. It's all there.
Doug's pot busses are actually grounded at the input jacks. Notice there are two pot busses. It was convenient and uncluttered to place that ground symbol at the right end of the buss. THE BUSS IS NOT CONNECTED TO CHASSIS AT THE POINT OF THE GROUND SYMBOL!
-
That is confusing. I wish Doug would move that ground symbol.
If you study Doug's layout you'll see exactly how he wired the shielded cables for the input jacks and the volume pots. Look closely. It's all there.
Yeah, I saw that the shields are grounded at the inputs, but I just wanted to confirm b/c of that ground symbol at the other end of the pre-amp ground bus. It's good to know that the pre-amp busses are grounded at the inputs. Thanks!
-
Started wiring today and ran into another question almost immediately: Robrob's layout shows the long horizontal wire runs (for example, between the V1B .047uF coupling cap on the far right of the board to the the V1 220K mix resistor on the far left of the board) as running under the board. But all the layouts I see (including the vintage Fenders) show these wires flown and twisted above the board. Is there any advantage to either method? Robrob's under-the-board seems much neater and I'd like to go with that, but I'm wondering if there's any advantage to flying/twisting the long wire runs on the board?
-
I can tell you that running the wires under the eyelet board such as the early blackface amps used is more work than flying them over the board. I think CBS Fender may have started running them overhead because it was easier and quicker thus saving production time. I don't know if there is any advantage of one over the other electrically. Is one quieter than the other?, I personally don't know about that but all the vintage Fender amps that I have worked on whether jumpers were wired over the board or under were very quiet amps.
-
Tidiness is one thing.. but you have to be 300% sure about the underboard wiring being done very well. It's near impossible to work on those afterwards when something turns out to be incorrect.
-
Especially when you're a beginner (anything less than 5 amps), the less you put under the board, the easier it is to troubleshoot and fix errors.
Tidiness is nice, but a working amp is king.
-
Thx for the quick responses! It sounds like the consensus is that there's no sonic difference, so I'm going to do the wiring under the board for neatness ala Robrob's layout. I just did two Tweeds with under-the-board wiring and had no issues (I'm pretty meticulous about wire routing) and I'll be careful as usual mounting the board.
-
Robrob's layout shows the long horizontal wire runs (for example, between the V1B .047uF coupling cap on the far right of the board to the the V1 220K mix resistor on the far left of the board) as running under the board.
Hoffman's layout, while quite different from the original layout, does the same for the wire in your example.
https://el34world.com/Hoffman/files/Hoffman_AB763_2.pdf
-
Tidiness is one thing.. but you have to be 300% sure about the underboard wiring being done very well. It's near impossible to work on those afterwards when something turns out to be incorrect.
It's not impossible,,, pull all the pots and jacks, unscrew the board mounts disconnect the HV leads on the PS side as necessary, and then lift the board and pots up to access. The brass plate may need some chassis soldered connections wicked up.. The underboard wiring lasted for decades without any problems ..
-
Tidiness is one thing.. but you have to be 300% sure about the underboard wiring being done very well. It's near impossible to work on those afterwards when something turns out to be incorrect.
It's not impossible,,, pull all the pots and jacks, unscrew the board mounts disconnect the HV leads on the PS side as necessary, and then lift the board and pots up to access. The brass plate may need some chassis soldered connections wicked up.. The underboard wiring lasted for decades without any problems ..
I said 'near' impossible... like you describe it's kind of a hassle to do this. I do as much on board jumpering as possible. The only wires that I might do under the board are ground wires, which I like to be as short as possible.
-
Making some progress here, and I'm noticing that some vintage Fenders and new builds use a "ground wrap" wire around the twisted wires coming from the doghouse (B+ 1, 2 & 3). Is this just a thinner wire with one end connected to ground and the other left unconnected? Does it make a difference, and to which ground point (pre-amp or power-amp) should it be connected?
-
The real blackface AB763 DR did not use a "wrap". Simply twist B1, B2, B3, and the ground connection for those caps together. The ground wire connects to the chassis at the power ground point. Look at these pics...
http://sluckeyamps.com/64DR/index.htm
-
OK, thx Slucky. Those pics answer a couple other questions, so thanks for the link.
-
Finished everything today and went through Robinette’s start up procedure step by step. I got to the final step with all tubes installed and the speaker plugged in. I get good plate voltages on the power tubes, but I get no sound from the speaker, and no “hiss” when the volumes are turned up. (I do get a very slight click when I probe the power tube grids.) So I stepped back to the previous step:. With the preamp and rectifier tubes installed (and no power tubes) I get correct B+ voltages, all the heaters work, and I get voltages on plates of all the preamp tubes EXCEPT the V3 (reverb driver). I checked all the wiring around this tube (also for continuity) and couldn’t find an error.
I should be seeing plate voltages on the V3 plates, correct? Could I have fried that 500pf reverb filter cap or the .022 coupling cap? The only other thing I can think of is the reverb transformer (photo attached). I remember when I purchased it it was for a Fender Pro, but I read somewhere it should work as a 125A20B replacement. Is this the correct transformer?
The only other anomaly is that I can’t get a voltage reading on pin 5 of V5. Shouldn’t there be a couple volts there w/out the power tubes installed?
(As usual, tried all the obvious stuff. The speaker works, power tubes are good, I’ve tried multiple sets of preamp tubes, and I’ve triple checked wiring every step of the way.)
Thanks in advance for any guidance you can provide!
-
Did you use Robinette's layout? If no, which layout did you use? If yes, read on...
V3 plates should have B+. Do you have B (B+2) where the red reverb transformer connects to the board? If not, follow the yellow wire to the doghouse.
V5 pin 5 is a filament pin. There should be about 3.15VAC on pin 5.
What voltages do you have on the 6V6 tubes, pins 3, 4, 5, and 8?
-
Did you use Robinette's layout? If no, which layout did you use? If yes, read on...
I followed Slucky's layout exactly, including wire colors. The only change I made was that I have pin 8 of the power tubes grounded at the socket through a 1 ohm resistor for easy bias measurements, rather than grounded at the power amp ground point.
V3 plates should have B+. Do you have B (B+2) where the red reverb transformer connects to the board? If not, follow the yellow wire to the doghouse.
Yes, I have B+2 of 459 at the red reverb transformer wire. Seems a bit high?
V5 pin 5 is a filament pin. There should be about 3.15VAC on pin 5.
Sorry, I meant V.5 pin 3 in my previous post, not pin 5. The heaters all work, but I only get a few millivolts on the cathode of the first V.5 triode.
What voltages do you have on the 6V6 tubes, pins 3, 4, 5, and 8?
3 = 462
4 = less than one volt if anything
5 = -36, good for the bias voltage?
8 = nothing, and this should be showing 30mv-ish, correct? As mentioned above, I have pin 8 connected to a 1 ohm, 3-watt resistor with the other side of this resistor connected to ground for easy bias measurements. I do NOT have pin 8 connected to the power amp ground bus as Robinette does. This is the only difference btwn. my wiring and Rob's layout.
Also, I noticed this morning is that I DO get a sound with the guitar plugged in and the volume all the way up, on all inputs. It's a quiet, distorted sound, like a distortion pedal with the battery dying. The volume must be maxed to hear any sound at all, and it's only the volume of someone talking.
The only other thing I was confused about was the labeling on those mylar caps around the tremolo (see photo). "F223K" = .02uf and "F103K" =.01uf, correct? I remember being confused about this when I purchased the caps, and I'm wondering if I installed the wrong values.
-
4 = less than one volt if anything
You have more than one issue but pin 4 of the 6V6s is the main issue. Fix it first. Look at Robinette's layout. There is a 470Ω resistor between pins 4 and 6 on the 6V6 sockets. A yellow wire connects pin 6 of the sockets together and another yellow wire connects to an eyelet in the top left corner of the board. The choke connects to this same eyelet and a jumper connects to the eyelet directly above it. This eyelet is labeled "B (B+2)" and has a yellow wire going to the doghouse. Make sure this is all correct. There should be about 460V at all these points. If the solution is not evident, open the doghouse.
-
You mean those two yellow wires THAT I COMPLETELY FORGOT TO INSTALL? Wires installed, crow digested. On the power tubes, I now get:
Pin 3: 460
4: 460
5: -60
8: nothing
Still no plate voltage on the V.3.
I get some distorted guitar audio, very loud hum, and very loud crackles and pops. (Is it OK to use a very old Jensen P12S with cone rub to test? Don't want to ruin my Celestion blue!)
Where should I go from here? And happy new year!
-
One down. More than one to go.
I think you should unplug and compare the schematic and layout to your actual build. Use a highlighter to mark the schematic and layout as you verify every wire and component value.
-
I think you should unplug and compare the schematic and layout to your actual build. Use a highlighter to mark the schematic and layout as you verify every wire and component value.
Did this before startup, and did it again this morning for 2 hours. I did forget to install those B+2 wires, but today I found no other wiring errors or errors in component values. The only components I don't know how to check are the Alpha pots, but I measured every pot before installing.
At this point, the only difference btwn. my amp and Rob’s layout (that I can find) is the white wire that goes from the 2.2m “Trem Jump Start” resistor to the bias board. Instead of connecting the wire to the bias board, I connected it to the bottom tab of the bias pot. No difference, correct?
I did find 2 issues today. The “1 ohm, 1%” resistors that I connected from pin 8 to ground turn out to be somewhere btwn. 5-7 ohms, so I removed them and hooked up the pin 8 grounds exactly like Rob's layout. I ordered these resistors last minute and they are the only ones I didn’t measure before installing.
I also found a bad cap that was a source of much of the noise. The .001 V.6 coupling cap hissed and popped when chopsticked. I didn’t have a replacement so I temporarily put 3 x .0047 caps in series as a temp. fix, and that solved a lot of the noise.
For the umpteenth time I replaced every preamp tube one by one. At this point, I can hear the guitar audio clearly but it distorts way too early and is not as loud/full as it should be. It kind of sounds as if I’m using a blown speaker. I notice two issues:
1. Still no plate voltage (.16v) on V.3. I’ve tried 3 tubes in that position and checked everything around it and I can’t solve it. Aside from randomly replacing components around V.3, I’m not sure where to go from here.
2. Something seems “off” with the bias adjustment. When I turn the bias pot I get all types of squeal and noise, and just stopped turning at a point where there’s no noise. Also, since I removed those 1 ohm resistors I don’t understand how to measure the bias. My plate voltage and OT CT voltage are the SAME (is this normal?), so I get zero when I subtract the plate voltage from the OT CT voltage.
Everything else looks good, and the Reverb and Trem seem to work perfectly.
Voltages:
B+1: 459
B+2: 457
B+3: 360
B+4: 300
6V6s:
1: -60 (Is this right?)
3: 459
4: 459
5: -60
6: 457
8: 0 (Pin 8 is now grounded as in Rob’s layout)
V1, 2, 4: plates are all around 190, cathodes are 1.5-3v
V3: .01v plates (!), 11v cathodes
V5: first plate is 441, second plate is 88, cathodes are almost 0
V6: 200ish on both plates, 86 on the cathodes
5 hours later I’m out of troubleshooting steps. Where do I go from here?
-
6V6s:
1: -60 (Is this right?)
NO! Adjust the bias pot to put -40V on pins 1 and 5.
Still no plate voltage (.16v) on V.3.
You said earlier that you had good voltage at the board eyelet the RT red wire connects. The RT blue wire that connects to V3 pins 1 and 6 should have about the same voltage on it. If not, the RT is likely bad. Use your ohm meter to check resistance between the RT red and blue wires. SHOULD NOT BE OPEN!
However, you have some conflicting numbers on V3. If you have only .16V on the plates, then it's impossible to have 11V on the cathodes.
-
I adjusted the bias pot so that there's about -40 volts on pins 1 of the 6v6s, but I get so many squeaks and skwawks at this setting that I'm afraid to damage my speaker and every measurement makes a really loud noise. There's a high-pitched whine coming through the speaker at this setting. I did have bits of snipped wire floating around when I was wiring. Tried to be careful and account for ever one, but I wonder if something got into the bias pot. I'll take it apart tomorrow or Mon and test it.
The resistance btwn. the red and blue wires on the reverb transformer is about 1.2K ("1.2" on my meter when I have it at the 20K setting.) The blue wire of the RT reads over 400v. I just measured it really quickly, but I think it's close to the red wire/B+2.
Sorry, I meant "1.1v" on the V.3 cathodes w/almost no plate volts, not "11".
-
The blue wire of the RT reads over 400v.
That blue wire should be connected to V3 pins 1 and 6 (plates). Yet you say there is no volts on the plates. What pins do you think are the plates???
Look at Robinette's layout. There is a yellow wire connected from the speaker jack to an 820Ω resistor on the board. Disconnect this wire (either end). Do the squeaks, squawks, and whines go away with the bias set for -40v at pin 5 of the 6V6s?
-
Sorry, I was mistakenly was reading white wires across the tubes. There's no voltage on V.2 2&7, not 1&6! I'll try that negative feedback wire.
-
So yes, the high pitched squeal goes away when I disconnect the neg. feedback loop. Should I change NFB resistor to another value? Still getting distorted audio and excessive speaker hiss...
-
I suggest to flip the power amp signal polarity, then reconnect the NFB.
eg swap over the wires at at OT primary or secondary, or at the output valve control grids.
-
… The “1 ohm, 1%” resistors that I connected from pin 8 to ground turn out to be somewhere btwn. 5-7 ohms, so I removed them and hooked up the pin 8 grounds exactly like Rob's layout…
What does your resistance meter read with the probes touching?
How do you now plan to check that the bias is ok?
-
So yes, the high pitched squeal goes away when I disconnect the neg. feedback loop. Should I change NFB resistor to another value? Still getting distorted audio and excessive speaker hiss...
Look at Robinette's layout. There is a blue OT wire connected to pin 3 of one 6V6 and there's a brown OT wire connected to pin 3 of the other 6V6. Swap those two wires and reconnect the NFB wire.
It's time for you to post some hi-rez closeup pics of your amp. I'm tired of working with a blindfold on.
-
Hi gf
Measure your meter probes, on "OHMS" but shorted. This becomes your "reference" for any immediate resistance measurements. Now measure your 1 ohm/1% resistor! It should be 1 ohm higher then your "ref" measurement. (Always repeat this when doing multimeter resistance measurements, so that you know "whereabouts in the paddock you are situated" ie referenced to something) This is valid whether you have a $10 cheapie or a $1000 Fluke.
The 1 ohm resistors are a very good tool for a quick check of you bias current. In my opinion a handy tool to simplify your amp building life.
Changing your NFB resistor will make little difference to the situation if your output tube wiring is out of phase.
We look forward to some hi res pics :icon_biggrin:
Regards
Mirek
-
or a $1000 Fluke.
:laugh: I'll sell you mine for 1/2 that :icon_biggrin:
a nice feature on the better meters is a zero button, short the leads, hit button, walla, display says zero.
-
Hi glass54 & pdf64, your right, they are 1ohm! Great tip. Now I just have to install them again...
Just to understand, if I install these from pin 8 to ground, I still need to run a ground wire from pin 8 to the power amp ground, correct? I had initially thought that these resistors would *replace* the ground wire.
Slucky, swapping those blue/brown wires worked! No more whine.
I'm still getting distorted audio and lots of "shhhhhhh" through the speaker. I'm going to check tubes and remeasure voltages ASAP.
Lots of photos attached.
-
More photos....
-
...I had initially thought that these resistors would *replace* the ground wire....
You were initially right. Why would you spend cash and time to buy and test resistors, and then run a wire AROUND them?
Why build a toll booth on the turnpike, and then build a wide-open bypass lane around it?
-
Ah, ok. So I'll remove that. Should I keep the wire that ties the 6v6 pin 8's together?
-
Should I keep the wire that ties the 6v6 pin 8's together?
no
-
...and you have the 1R fitted correctly (as you did originally)?
You can let us know what your bias current is in each o/p tube
:laugh:
-
Yep, sorted! 22mv across the 1ohm resistors, so 22ma-ish on each 6v6. The tone is much stronger now, and it appears the only issue is the excessive "shhhhhh". What could cause this? I notice that when I push down on V.6 B Grid Leak eyelet the "shhh" gets slightly quieter. I disconnected that leg and tried another cap, but it doesn't get better. It does seem like the noise is coming from that area tho....
-
All those 1/2 watt carbon comp resistors are a huge contributor to shhh.
Divide and conquer. Pull V4 Does the shhh go away?
-
...and, it's finished! Huge, full, shimmery sound through my 5e3's Celestion blue, and the reverb and trem both work great! The loud "shhhh" is the reverb when the tank isn't hooked up and the reverb control was maxed out. (I don't mind the mild "shhh" from CC resistors. My fave amps have them.) The reverb seems a bit noisy and the trem clicks a bit, but no more than any other DR I've owned. After the biasing, the voltages are about 5-10% lower than those posted above.
So now I have to order that coupling cap, and I'm going to redo the power tube sockets that are a mess. Waiting for a speaker from Weber and I'm hoping to order a Mather cab when I get some extra $$.
I can't say how much I appreciate all the help, and if it wasn't for you guys (especially Willabe and Slucky) these amps would be half-finished gathering dust in the corner. I now have a fantastic sounding 5E3 ("Alpha"), a 5F6A Tweed Bassman ("Delta"), and this DR ("Omicron") thanks to this forum. That said, I think this is gonna be it for me. Tomorrow I go back to being a musician and I'll leave this stuff for the experts.
-
Well done gf :thumbsup: Very neat job.
....But why stop building now that you have developed some very good expertise :occasion14:
Kind regards
Mirek