Jump to content

Eico HF-87 Clone Build


Curious_George

Recommended Posts

3 hours ago, Curious_George said:

I had the wrong cap selected for use in the formula...

 

Maybe I used the wrong cap value? I can't remember what value you (or I) used for the formula as I don't see it in any of the posts. Since you are at 15Hz for resonant frequency now you must have around 220uF in there now? Most of the text books on power supplies for audio amplifiers using LC filters states the resonant frequency should be ideally lower than 10Hz but they might be shooting for flat response down to 20Hz. Of course the high pass filters for the coupling stages could be set high enough to not cause issues with the resonant frequency. Same goes for if the output transformer or anything else limiting low frequencies etc...

 

If you are using 20kHz square waves that is like testing for 200kHz, I usually start at 1kHz and work my way up to 20kHz. There were some instances where 10kHz was worse than 20kHz, I'm thinking it depends where the resonant frequency is at in the output transformer? That was odd which is why I just test at different frequencies in case there is an anomaly.

 

The amp looks great underneath and it sounds like it's working great for you. How much feedback are you using currently?

 

 

 

Link to comment
Share on other sites

22 minutes ago, Curious_George said:

15.5dB. I'm sure I have headroom to use more, but I am happy with the sound and the distortion figures with 15.5dB. 

 

That's the great thing about feedback, use as little or as much as you like. You have local feedback in your output stage with the distributed loading which is usually the worst offender of distortion so less global feedback makes sense. Since I typically build Pentode operation I have found that 30db or more of feedback gets tough to keep stable with off the shelf EI core output transformers. I found that 20db or slightly higher if i have enough gain is good amount to improve performance but still end up with a healthy phase margin after compensation to where square waves look good even with capacitive loads.

 

 

Which brings me to my next question. Have you added any reactance to your bench load with the square wave testing? That's how I find if an amp is marginally stable without having to work out it's actual phase and gain margins. In the real world with loudspeakers as a load can push some amps into ringing where on the bench it looks fine. Which is partly why even if I still get decent square waves with the loop closed I continue on to adding compensation networks for the sake of increasing phase margin past the point where I feel confident the amp won't ring with any load it's hooked up to, even electrostatics. Heck I'll even just hook a capacitor up to the amp as the load to see what happens. I'd feel bad if someone hooked my amp up to some ES panels and it at best produced poor sound or worse, damage from the amplifier turned oscillator. I cloned a popular guitar amp load box which contains all passives to emulate a voice coil electrically. I use that sometimes as a "speaker load" test but it's only good for 50 watts. Most of the time I just use my 1800 watt dummy load and hook caps up across it, works just fine.

 

So if your amp keeps stable on a capacitive load then you should be good to go. With 15db of feedback with that circuit you may find it doesn't like the capacitance which is where adding the compensation networks will help. From memory with Edcor transformers a 5-20 circuit with that much feedback will most likely start to misbehave with some reactance. Well, at least for my amps they did but I don't remember testing exactly 15db, more closer to 20db. I have a simulation file of the UL amps I made a while back so all I have to do is go in and pop in the resistor values I was messing with (3.3k-6.8k me thinks) to see how much feedback I got at minimum, which would be a 6.8k resistor. That's with using the 100 ohm resistor for the divider network at the input stage cathode.

 

 

 

Link to comment
Share on other sites

4 hours ago, Curious_George said:

Although I don't have any issues with the current power supply components or values of components, I went ahead and added some extra capacitance mainly because I had room for the capacitors. In addition, I am using an NTC in-rush current limiter (CL-90) so this will take care of any start-up current issues with the extra capacitance. 

 

The solid-state power supply helps maintain a low impedance which is necessary for fast tight bass and overall performance. "They" wish they had large inexpensive electrolytics back in the 50's like we do today. 

 

With the added capacitance the new resonant frequency of the PS is now 15.18Hz. 

Eico HF-87 Clone Bottom_More Capacitance.jpg

Looks really nice and neat. I like. 🙂

Link to comment
Share on other sites

Nick is one of the cleanest builders I have seen on here. I love that sort of stuff.

 

I apparently catch flack for taking too long to build amps but I'm very anal about things. I like the coupling caps going to the tag boards like you did there to keep them from moving around, it's hard to see but a lot of those radial caps don't like the leads bent coming right out of the body. It breaks the seal or "meniscus" and moisture may creep inside. I'll post some images and the explanation for a stress loop except you aren't going down to a board, bring the leads to the tag boards. A little dab of resin/epoxy where the cap meets the chassis helps with end mounted parts.

 

The leads are thick enough to still hold the capacitor stiff against the chassis. Nice rugged layout. I see a lot of builders skip the tag boards and just solder one leg to the tube socket and bring a wire to the other leg, the capacitor kinda just floats around by it's leads and the wire. Not very professional looking in my book. But I'm anal probably because I was programmed to meet a certain standard because if something failed it's would be your butt on the line. That and I always just loved to see rugged MIL spec electronics, it's a science that doubles as a art form almost. Some of those Japanese guys are fantastic builders, you could tumble the amp down a hillside and it would still function perfectly. Just amazing.

stressloop.png

stressloopexp.png

Link to comment
Share on other sites

6 hours ago, captainbeefheart said:

If you are using 20kHz square waves that is like testing for 200kHz, I usually start at 1kHz and work my way up to 20kHz. There were some instances where 10kHz was worse than 20kHz, I'm thinking it depends where the resonant frequency is at in the output transformer? That was odd which is why I just test at different frequencies in case there is an anomaly.

I’ll do a more thorough test regimen. I was so excited to get this amp built and up and running. I’ve been waiting a year to build it. 

Link to comment
Share on other sites

4 hours ago, captainbeefheart said:

Which brings me to my next question. Have you added any reactance to your bench load with the square wave testing? That's how I find if an amp is marginally stable without having to work out it's actual phase and gain margins. In the real world with loudspeakers as a load can push some amps into ringing where on the bench it looks fine. Which is partly why even if I still get decent square waves with the loop closed I continue on to adding compensation networks for the sake of increasing phase margin past the point where I feel confident the amp won't ring with any load it's hooked up to, even electrostatics. Heck I'll even just hook a capacitor up to the amp as the load to see what happens. I'd feel bad if someone hooked my amp up to some ES panels and it at best produced poor sound or worse, damage from the amplifier turned oscillator. I cloned a popular guitar amp load box which contains all passives to emulate a voice coil electrically. I use that sometimes as a "speaker load" test but it's only good for 50 watts. Most of the time I just use my 1800 watt dummy load and hook caps up across it, works just fine.

 

So if your amp keeps stable on a capacitive load then you should be good to go. With 15db of feedback with that circuit you may find it doesn't like the capacitance which is where adding the compensation networks will help. From memory with Edcor transformers a 5-20 circuit with that much feedback will most likely start to misbehave with some reactance. Well, at least for my amps they did but I don't remember testing exactly 15db, more closer to 20db. I have a simulation file of the UL amps I made a while back so all I have to do is go in and pop in the resistor values I was messing with (3.3k-6.8k me thinks) to see how much feedback I got at minimum, which would be a 6.8k resistor. That's with using the 100 ohm resistor for the divider network at the input stage cathode.

I don’t have any reactance loads per say, just resistors. But I have been wanting to build a reactive load for testing. It’s on my project list. 

Link to comment
Share on other sites

2 hours ago, captainbeefheart said:

Nick is one of the cleanest builders I have seen on here. I love that sort of stuff.

 

I apparently catch flack for taking too long to build amps but I'm very anal about things. I like the coupling caps going to the tag boards like you did there to keep them from moving around, it's hard to see but a lot of those radial caps don't like the leads bent coming right out of the body. It breaks the seal or "meniscus" and moisture may creep inside. I'll post some images and the explanation for a stress loop except you aren't going down to a board, bring the leads to the tag boards. A little dab of resin/epoxy where the cap meets the chassis helps with end mounted parts.

 

The leads are thick enough to still hold the capacitor stiff against the chassis. Nice rugged layout. I see a lot of builders skip the tag boards and just solder one leg to the tube socket and bring a wire to the other leg, the capacitor kinda just floats around by it's leads and the wire. Not very professional looking in my book. But I'm anal probably because I was programmed to meet a certain standard because if something failed it's would be your butt on the line. That and I always just loved to see rugged MIL spec electronics, it's a science that doubles as a art form almost. Some of those Japanese guys are fantastic builders, you could tumble the amp down a hillside and it would still function perfectly. Just amazing. 

 

The new Orange Drops I used in the HF-87C are different than ones I’ve seen in the past. The orange epoxy coating seems to be a different composition too. The epoxy coating around the leads is now “machined” clean around the entire lead out area. Previous Orange Drops had the epoxy coating free flowing around the leads. This made bending the leads a challenge as you mentioned because small pieces of the epoxy would break off and not in a clean manner. I’ll get a close up pic of the new Orange Drops with the leads.

 

The new Orange Drops are much easier to work with and they look clean. 

 


 

 

Link to comment
Share on other sites

7 hours ago, Curious_George said:

High voltage, filament, input, driver, output and feedback. 

I usually just figure the active circuits, the ones around the tubes. I see 7 tubes. 

 

I have an Eico ST-70 I started modifying and restoring a number of years ago and put on hold. This week I tested the tubes in it at Classic Audio and to my surprise all but 2 of them tested as new. The other 2 tested 97%. Not having a tube tester and fearing some of them may be bad was one reason I put the project on hold. Now I know I have really good iron and good tubes I am thinking why not just use these parts and build an amplifier using one of my wood chassis with less controls and frills instead of restoring it. The tubes are all original ones and them and the iron are worth more than I paid for the entire amp years ago. Below is the link on this site when I bought this amp. I was in the process doing the Tranola mods Mike Stehr mentions when I put it on the shelf. 

https://community.klipsch.com/index.php?/topic/196477-eico-st-70/

 

Link to comment
Share on other sites

22 hours ago, captainbeefheart said:

I like the coupling caps going to the tag boards like you did there to keep them from moving around, it's hard to see but a lot of those radial caps don't like the leads bent coming right out of the body. It breaks the seal or "meniscus" and moisture may creep inside. I'll post some images and the explanation for a stress loop except you aren't going down to a board, bring the leads to the tag boards. A little dab of resin/epoxy where the cap meets the chassis helps with end mounted parts.

 

The leads are thick enough to still hold the capacitor stiff against the chassis. Nice rugged layout. I see a lot of builders skip the tag boards and just solder one leg to the tube socket and bring a wire to the other leg, the capacitor kinda just floats around by it's leads and the wire. Not very professional looking in my book. But I'm anal probably because I was programmed to meet a certain standard because if something failed it's would be your butt on the line. That and I always just loved to see rugged MIL spec electronics, it's a science that doubles as a art form almost. Some of those Japanese guys are fantastic builders, you could tumble the amp down a hillside and it would still function perfectly. Just amazing.

Here are a few close-ups of the Orange Drops. I was going to use Orange Drops too for the grounded grid half of the phase inverter, but the radial leads did not lend themselves to a nice clean layout, so I used Illinois Capacitor (IC) metallized film (polyester) caps in those locations. 

Orange Drop Cap_1.jpg

  • Like 2
Link to comment
Share on other sites

24 minutes ago, henry4841 said:

So you like those orange drops as I do. Sound good to me as well. 

I do. They are one of the few film & foil caps that are nominally priced and have a long history of use. I still love metallized film, but for the price, it is hard to not use an Orange Drop for signal coupling.  

 

"Drops of Jupiter" are priced out of this world. 

  • Like 1
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...