You can find more information on our web, so please take a look.
This post is about mainly sharing the experience of a small maker business sourcing genuine components right here in the USA. Sourcing and stocking the right components in the right quantities is one of the finest balancing acts in a small electronic manufacturing business. That&#;s because keeping large stocks of parts has not only tax implications but also drains funds from other efforts faster than you may realize. If you&#;re a beginner and need some tips, here&#;s some experience that might save you a little sweat and hopefully a few of your hard earned dollars. I might will update this article in the future as more things come to be realized in this area of the business.
My main parts suppliers since LowPowerLab&#;s inception were Mouser and Digikey. Both are great in their ways which I outline down below, but there have been some surprising downside quirks as well. Hey it&#;s a little hard to get motivated to write only about awesome things, usually the bad stuff triggers reviews. But I&#;ll try to keep this objective and informative. If both were great at everything then there would be no reason for competition or even having multiple suppliers. Monopoly is a bad thing in supply chains, and I hope these companies will continue to exist as separate entities and not eat each other in some monster acquisition deal.
I have mainly 2 types of orders. One type is the &#;small value&#; prototyping batch of components or small run custom batch for a client which I usually place at Digikey due to their more comprehensive offering. Even this can run into the lower hundred dollar range depending on the project I&#;m putting together, but the average is under $100. Then there is my normal run, what I call the periodic replenishing orders which are in the thousands of dollars range. Whenever I run low on certain parts, I place them in the cart, and they stay there until I&#;m ready to complete the order. It&#;s unlikely for me to use fast shipping, so I usually go with ground services. I even use USPS for the real low count prototype orders.
Before I get into my sourcing experience with these suppliers, let&#;s take a quick look at the pretty incredible state of affairs with respect to passive component sourcing in Q1/. So as a quick primer side story about parts sourcing &#; if you&#;re a maker you&#;re already familiar with the astronomical passives lead-time situation. Just the other day I was scrambling to find 0.1uF caps (which along with 1K 10K resistors are the most popular and numerous parts in my designs). They nowhere to be found these days due to extreme factory lead times for new passive stocks. I ended up paying $210 this week for the last 15K reel of cheapest 0.1uF caps (normally <$20 for a 4K reel). Yup, it&#;s that bad!
Neither of these companies are directly at fault for this, at least not for the supply chain. But they control their dwindling stocks pricing of course that&#;s for sure, and why not make a buck while in this dire straits, right? This week I made the mistake of keeping some reels of basic passives in the cart for a day or so until I completed my order (at Digikey first then at Mouser when Digikey 1) doubled the price then 2) went from a comfortable 16-24,000 parts to 0 stock. No problem, Mouser doesn&#;t mess with pricing right? Then Mouser did it too, but too late for screenshots to prove it! I should have just bought those on the spot knowing everyone else is trying to buy them too. Don&#;t procrastinate silly!
Anyway let&#;s now go ahead and imagine a little fable here:
In general as companies (not necessarily these) get larger they tend to become more imperialistic, trusting their dominance on market share confidence and getting preferential with customers depending on their bottom line impact. Small customer care becomes secondary. Me and you are probably lost in the background noise of annoying small customers, more so the lower the order value (more overhead per order for less profit!). Say you bought a few different SMD resistors, some switches, whatever else small value components for your latest breadboard project.
You buy not-overnight shipping, place your order sometime in the morning and expect them shipped in cut tape or whatever part spec packaging, before the same-day shipping 8-9PM cutoff time, well in time to get it for your exciting weekend project you&#;ve been looking forward to. But hey wait &#; who cares about your $19 order value, when there&#;s real customers. Think someone will walk the warehouse 5 minutes to count and cut your 10x resistor tape then pack your $19 order by cutoff time just because your order came in early? Have a question or concern? Go away annoying fruit fly, we are having some difficulties fulfilling the unexpectedly large number of orders in the last day, we will get to your order after the real customers and our interests get served.
That fable is a little on the extreme. But it get&#;s pretty close to reality actually when it feels like it&#;s happening to you, again, then again, and again.
Let&#;s look at the pros and cons which I&#;ve seen in the past few years of dealing with these suppliers.
The goods:
The bads:
One of my main components is the venerable atmega328p which I buy in trays of 250x, usually at least 2-4 trays each time. There are expensive $3-6 parts/sensors which I buy as a partial digireel simply because keeping $ invested in a full reel is a little scary to me (what if it goes obsolete and the arduino world moves to the new version like it happend with the BMP180 and newer BME280 sensors). When I buy full trays or more than x of any taped component, I expect a reasonable effort to be served full trays and 1 continuous tape of that component rather than several pieces of tape. It&#;s common sense that these items will be machine placed, not hand placed.
Digikey&#;s dodgy packaging
Here&#;s a real repeating scenario. Digikey has been consistently dropping the ball on component packaging and has been sending me &#;scrap&#; parts. Instead of a package of 2 full trays (500x), I&#;ve been getting a full tray + a partial tray + scrap pieces of cut-taped atmegas, packaged separately. So I have to do their job moving parts from tape to trays for use in the pick and place. In doing so I noticed various different date marking on the chips, so they are basically cleaning their parts bins and just dumping it to no-name customer me. Thanks Digikey, your time savings and inventorial efficiency remind me of how much I mean to you every time I sit there with tweezers trying to place each chip in the tray in the right orientation, and flat to prevent pin bending. What a pain. On the flip side, on a few occasions Digikey has sent me a few extra chips (the junk on the bottom of the bins? to make me feel good?). Thanks but I&#;d rather just have what I paid for (complete full trays) than $20 extra worth of their scrap parts!
Another woe with Digikey packaging is digireels sometimes come in 2-3 or sevearl split tapes which are poorly joined together. They almost always fail in the pick and place, the pullback tape just flies off, or the metal strip joining piece rips apart, its just junk. I basically never order digi reels anymore either &#; their leader tape is just as useless, a waste of extra $7 digireel fee. I just pray every time that they will send me 1 continuous piece of tape and I deal with the leader tape myself.
On several occasions Digikey sent me tray parts that have bent pins (improper stacking of trays or chips were out of place and stacking trays bends the pins), this of course goes unnoticed until testing, and requires manual rework to salvage the failing boards. This type of thing is only fun while you read about it. Meanwhile Mouser never managed to match this performance:
Mouser&#;s packaging
Mouser on the other hand has are stellar at extra care with packaging and padding the trays before they package them. And just as great with dry packs and professionally packet parts. I don&#;t recall ever being upset with them cleaning their component bins and dumping N number of split tape of a certain component, it&#;s always nice to have 1 continuous piece of tape for use in the pick and place. No complaints, just great service.
Now part two of Digikey&#;s mishaps. Getting my order shipped the same day at Digikey has been hit or miss. This last last one (placed April 12, 10am) shipped the next day (April 13 12pm) using the same ground service. So their same-day shipping falls short and I have to eat any potential delay penalties myself (or just spend a lot more on shipping in hope they might ship the same day). The order was just over $ so maybe a little off their priority radar, oh well!
Enter Mouser. At the same time (minutes apart) I placed $500 order, also with ground shipping. Surprisingly (I can&#;t recall them ever doing this) they also delayed shipping past the promised same-day timeframe. Very unusual. However, I got a nice saying my shipping was upgraded because of the delay, to Next-day Air. Nice, now I&#;m getting my parts even faster so the delay actually makes things better.
And as a past experience with Mouser &#; it seems like if you spend at least $500- they always upgraded me to next day air, sometimes 2nd day air, it&#;s pretty much routine. I don&#;t mind that, thanks Mouser, you have been awesome, please don&#;t change!
I&#;ve had good experience with:
Newark/Element14 seem to be pretty disabled in terms of stocks and shipping to the USA, I think I ordered once or twice, pretty helpless sites and part search engines.
Arrow &#; I will avoid it in the future. I ordered from them a few times, my last order was a pretty incredible experience buying a large batch of components. I was in a crunch and had no other choice but at Digikey who charged double, so I bit the bullet. The site is pretty helpless, parts ship from various global locations and this last order took &#;forever&#; to arrive from Netherlands. A few months later some rep Jack K. starts calling me to pay them some extra due tax they added after I placed the order (never mentioned at checkout or on the original invoice) &#; yeah this really happened in . I paid just to get them off crawling on my back, now I get Jack&#;s spam every week or two trying to &#;bid&#; me on more parts. Dear Arrow and Jack &#; you don&#;t charge a customer again after the deal &#; you and your stone age practices are dead to me and you lost my business for eternity.
Made in china?
So dear reader &#; has this article been helpful to you?
Have you had similar experiences with these or other suppliers?
Leave a comment, even hate mail is OK if I can learn from it, suggest improvements, share your own woes or awesome tips, anything constructive is welcome!
Back to work! Ready to stuff some boards?
Fundamentally, Schottky diodes and fast recovery diodes are not the same in physical structure. The anode of Schottky diode is metal, the cathode is N-type semiconductor. The basic structure of the fast recovery diode is still an ordinary PIN diode, that is, the anode is N type semiconductors and the cathode is P type semiconductors. The physical structure determines the electrical properties of both.
1. The voltage resistance is low (usually below 150V). With the same voltage resistance and the same current, the forward voltage of the Schottky diode is lower than the fast recovery diode. According to the figure, when under 1A current, SS15 has a Vf of about 0.65V and ES1A has a Vf of 0.85V under 1A (both could withstand the same 50 voltage).
2. The Schottky diode carrier has only electrons. And the multi-digit device theoretically has no reverse recovery time. While the fast recovery diode is essentially the same as the PIN diode. And the reverse recovery time of the low-digit device is usually in the tens to hundreds of ns. What exactly is reverse recovery will be discussed in detail in Question 2 of Post 2.
3. Under the rated reverse voltage, the reverse leakage current of the fast recovery diode is small, usually in a few uA to dozens of uA. The reverse leakage current of Schottky diodes usually reaches hundreds of uA to tens of mA, and increases sharply with the increase of temperature.
The basic definition of reverse recovery time is: the time required for the diode to transition from the on-state to the off state.
It can be seen from the definition that a reverse bias voltage is suddenly applied in the diode conduction state, and it cannot be cut off immediately. Let's figure out the reasons basically.
The picture above is the most basic structure of our most commonly used PN junction diode. The heavily doped P-type semiconductor acts as the anode and the heavily doped N-type semiconductor acts as the cathode. In order to withstand higher reverse voltage, there is a long region in the middle for the light doped N-type region close to the intrinsic semiconductor.
When PN junction is added with positive bias voltage, under the action of electric field, the holes in P region diffuse to N region, and the electrons in N region diffuse to P region, thus forming a forward current. For the P region, the electrons from the N region are minority. And for the N region, the holes from the P region are minority. Therefore, the PN junction diode actually transmits current through a small number of particles, that is, a small number of particles device. It can be seen that when the positive pilot passes, there are a large number of electrons in the P region and a large number of holes in the N region. The following take CCM's boost as an example, combined with simulation to analyze the diode reverse recovery problem.
The simulation results of the next steady state cycle are shown below:
V002 is the Vgs of the MOSFET. V(N003,N002) is the Vka of the diode. Id (M1) is the drain current of the MOSFET, and I (D1) is the anode inflow current of the diode.
When the new switching cycle comes, the MOSFET drive voltage rises. The inductive current begins to flow from the MOSFET, because the inductive current is basically unchanged in a short time. The current that the diode begins to flow begins to decrease. The moment the diode current decreases to 0 is the moment when reverse recovery begins.
Hornby Electronic Product Page
Due to the few minority stored in the diode, the diode cannot block the current. As the MOSFET drive voltage rises, the loop impedance between MOSFET DS decreases. And the diode begins to reverse the current flow, that is, the form of the minority and reverse phase current is extracted until the voltage between MOSFET DS begins to drop (begins to enter the Miller platform). That is, when the diode starting voltage begins to reverse bias, the reverse current reaches the maximum value. After that, the reverse current begins to decrease and the diode reverse voltage begins to increase until the MOSFET Vds drops to 0. And the diode withstands the full reverse voltage.
It can be seen from the simulation diagram that in the process of reverse recovery, the diode flows through a large reverse current and bears a large reverse voltage, so a large reverse recovery loss is caused. In CCM PFCS, in order to reduce the loss, the usual ultra-fast recovery diode (nominal reverse recovery time of ten to tens of ns) is still unsatisfactory, and the SiC diode is needed. The commonly used SiC diodes are usually Schottky structures, and the reverse recovery time is much lower than that of PIN diodes.
The answer is average current. (The subject being discussed is the on-state loss of diode. And the reverse recovery loss is not considered.) Why is that?
First of all, the basic characteristic of the diode is a single direction channel, so the current is one-way (do not consider reverse recovery), and the diode on-voltage drop is generally considered to be basically unchanged. Obviously, the product of the current and voltage within a certain time, the result is the product of the average current and forward voltage drop, that is, the diode on-state loss is calculated by the average current.
Is the effective current meaningless? Not really. In addition to the PN junction, the diode also has the contact resistance and lead resistance of the package. The loss of these parts is evaluated by the effective value. But relative to the junction voltage drop, the voltage drop brought by the package can be ignored in most cases.
The maximum operating temperature is determined more by the packaging material and the packaging process than by the semiconductor itself. For Si materials, 175 degrees is usually no problem. Temperature resistance of SiC and GaN are higher. The same voltage resistance, the same current diode, the wafer area PIN occupy is much smaller than the Schottky diode.
About the voltage resistance of diodes:
Schottky's voltage is usually up to 200V (Taiwan Zhumao and PFC brands said it could be 300V, and the VF value could be 0.3). In fact, the reverse recovery time of Schottky is fast because its chip is very thin (the current is related to the size of the wafer area, and the voltage is related to the thickness of the wafer), and the distance between A and B is short. The ultra-fast recovery (SF,MUR,ER series) can withstand 600V, and the fast recovery (FR,HER,RS series) can withstand V.
The maximum voltage of Schottky could be higher than 200V. But the silicon material restricts it. When the voltage is above 200V, Vf has no advantage relative to fast recovery. And the reverse leakage current is very large. A Schottky diode that can withstand 300V and Vf 0.3V has not been heard of and is theoretically unlikely. Schottky has no reverse recovery because the carriers are all electrons and there is no time to extract the minority.
The rated current of the diode is the main nominal value of the diode. Take 5A/100V diode as an example, 5A is the rated current. The rated current is usually defined as the average rated current through which the diode can pass. However, some tests are preceded by square waves, that is, square-wave currents that can pass an average of 5A. Some tests are based on DC. In theory, for silicon diodes, the diode with square wave as the test condition can pass through a larger DC current. Because the square wave of the same average current will bring more losses to the diode than the DC current. Is the diode of 5A able to pass through the current of 5A? Not necessarily. This is related to temperature. When your heat dissipation conditions are not good enough, the current that the diode can pass through will be limited by the junction temperature.
When the diode is reversed, it can withstand a certain amount of back pressure. The highest back pressure it can withstand is the rated voltage. For example, 5A/100V diode, its rated back voltage is 100V. Although, all diode manufacturers will leave a certain margin. 100V diodes usually use 110V will not have a problem. However, it is not recommended. Because more than the rated value, the manufacturer will not guarantee its reliability. And many power supply design companies, in order to ensure reliability, will also reduce the rated value.
When switching power supply in running state or in other transient situation, the diode needs to be able to withstand a large impact current. Certainly, this impact current should be not repetitive.And the interval should not be too long. The impulse current is usually defined in the data manual of the diode. And the test condition is often a single waveform of the impulse current, such as a single sine wave, or square wave. Its current value could be up to several hundred.
When the diode works, it creates a voltage drop. This voltage drop is related to the forward current and temperature. As for silicon diodes, the greater the current, the greater the voltage drop. The higher the temperature, the lower the pressure drop. However, the higher the temperature of the silicon carbide diode, the greater the pressure drop.
When the diode is reversed, it is not a completely ideal cutoff. When it is subjected to reverse voltage, some small current will leak from the cathode to the anode. This current is usually small. However, the higher the back voltage, the greater the leakage current. And the higher the temperature, the greater the leakage current. Large leakage current will bring large losses, especially in high voltage applications.
This is an important indicator of the diode. Fast or slow is based on the following standard. When the diode is converted from positive bias to reverse bias, a large reverse recovery current will flow from the cathode to the anode. And the reverse current will first rise to the peak and then fall to zero. The time of its rise and fall is the reverse recovery time. The peak current is the reverse recovery current. This can cause a lot of loss in high frequency applications. The reverse recovery time is positively correlated with the decline rate of the forward current when the diode is cut off. To solve this problem, one could use a diode with a faster recovery time or use the ZCS mode to turn off the diode.
When the switch converts from the on-state to the off-state, the diode or rectifier needs to first release the stored charge before the diode blocks the reverse current. This discharge time is called the reverse recovery time, during which the current flows through the diode in the reverse direction. That is, from the time when the positive pilot current is 0 to the time when it enters the full cutoff state.
The reverse recovery process is actually caused by the charge storage effect. And the reverse recovery time is the time required for the charge stored in the PN junction to be exhausted during the forward conduction. If a continuous PWM wave with frequency T1 passes through the diode, and when Trr < t1, the reversed diode could not blocked the PWM wave. And the switching effect cannot be achieved. Reverse recovery time of diode is provided by datasheet. The fast reverse recovery time enables the diode to quickly convert between on-state and off-state, which can obtain a higher switching speed, increasing the usage frequency of the device and improve the waveform.
Fast recovery diode features
The most important feature of the fast recovery diode is that its reverse recovery time (trr) is below a few hundred nanoseconds (ns), and the ultra-fast recovery diode can even reach tens of nanoseconds. The reverse recovery time (trr) is defined as a time interval (current through the zero from the forward to reverse conversion, and then from the reverse conversion to the specified low value). It is an important technical index to measure the performance of high-frequency continuous current and rectifier devices.
In the figure, IF is the forward current. IRM is the maximum reverse recovery current. And Irr is the reverse recovery current. Generally, Irr=0.1IRM. When t&#;t0, the forward current I=IF. When t > t0, because the forward voltage on the rectifier tube suddenly becomes the reverse voltage, the forward current rapidly decreases. And at the time t=t1, I=0. Then the reverse current IR on the rectifier tube increases gradually. The maximum reverse recovery current IRM value is reached at t=t2. After that, under the action of forward voltage, the reverse current gradually decreases and reaches the specified value Irr when t=t3. The reverse recovery process from t2 to t3 is similar to the discharge process of a capacitor. The interval from t1 to t3 is the reverse recovery time trr.
When the diode recovers in reverse, the reverse current drops relatively slowly, which is called the soft recovery diode. Soft recovery has some advantages in reducing EMI.
The junction capacitance is a parasitic parameter of the diode and can be regarded as a capacitor in parallel on the diode.
The parasitic inductance of the diode is mainly caused by the lead and can be regarded as the inductance in series on the diode.
For the transient process of diodes, the reverse recovery characteristic is usually the focus. However, in fact, the process of the diode from the reverse deflection to the positive pilot also has noteworthy aspects. When the diode switches on, the forward voltage drop will first rise to a maximum value, and then it will drop to the steady state value. And this maximum value increases as di/dt increases. That is to say, the diode band will generate a positive peak voltage at the moment of conduction. And the voltage is greater than the steady state voltage. The forward peak voltage of the fast recovery tube is relatively small, but the slow recovery tube is huge. Which brings us to another question.
RCD circuits are often used in situations where clamping is required, such as voltage clamping for flyback primary MOS and voltage clamping for secondary rectifier tubes. Some technical literature says that a slow recovery tube should be used, because the slow recovery tube has a long reverse recovery time, so that part of the energy in the clamp capacitor will be fed back to the circuit during the reverse recovery of the diode. And the loss of the entire RCD circuit can be reduced. But this is only suitable for small current or low di/dt like the low-power flyback&#;s primary side clamping circuit. However, it is not suitable for high-current, high-DI /dt clamping, such as the secondary clamping circuit of the power supply with a large current output. Because the slow recovery tube will produce a high on-voltage drop spike when it is on, resulting in a low voltage on the clamping capacitor, but it is unable to clamp the peak voltage. So we should choose something like a Schottky diode.
Usually, silicon is used as raw material of the diodes, but recently the more popular raw materials
Is silicon carbide. At present, the most common high-voltage Schottky silicon carbide diodes have the advantages of good reverse recovery characteristics, comparable to Schottky silicon diodes. However, diodes that can do high voltage have many applications in PFC. And its disadvantage is that the forward guide pass pressure drop is relatively large. Another difference with silicon diodes is that their on-pressure drop increases with temperature. The early silicon carbide diodes also have the disadvantages of small impulse current and low reliability. But there has been great improvement.
To be honest, I heard that gallium arsenide material was earlier than silicon carbide. But later I heard less about it. At present, gallium arsenide seems to have some applications in LED, but it is still relatively small on power devices.
Theoretically speaking, silicon diodes, because the on-voltage drop decreases with the temperature rise, are not suitable for parallel connection. But now many diodes will package two single tubes together, so that the temperature rise is relatively uniform, which brings benefits to parallel connection. However, the pressure drop of silicon carbide is increased with the rise of temperature. So it is theoretically suitable for parallel operation.
Get more information, please click the following link &#;https://www.aipulnion-power.com/index.php?m=home&c=View&a=index&aid=&admin_id=2
Contact us to discuss your requirements of China Schottky Rectifier Supplier. Our experienced sales team can help you identify the options that best suit your needs.